def clean_list_record(self, record):
if record['lat'] == '0' or record['long'] == '0':
raise SkipRecord('Got 0 for lat/long')
if record['lat'] is None or record['long'] is None:
raise SkipRecord('Got no value for lat/long')
if (record['lat'] == ''
or record['long'] == 0) and record['siteaddress'] == '':
raise SkipRecord('No value found for lat/long or address')
record['lat'] = float(record['lat'])
record['long'] = float(record['long'])
if record['servicetypecode'] in ('METERS', ):
raise SkipRecord('Skipping parking meter data')
record['order_date'] = parse_date(
record['serviceorderdate'].split('T')[0], '%Y-%m-%d')
record['add_date'] = parse_date(record['adddate'].split('T')[0],
'%Y-%m-%d')
if record['resolutiondate']:
record['resolution_date'] = parse_date(
record['resolutiondate'].split('T')[0], '%Y-%m-%d')
else:
record['resolution_date'] = None
if record['inspectiondate']:
record['inspection_date'] = parse_date(
record['inspectiondate'].split('T')[0], '%Y-%m-%d')
else:
record['inspection_date'] = None
if record['serviceduedate']:
record['service_due_date'] = parse_date(
record['serviceduedate'].split('T')[0], '%Y-%m-%d')
else:
record['service_due_date'] = None
record['service_notes'] = record['servicenotes'] or ''
record['inspection_complete'] = (record['inspectionflag'] == 'Y')
return record
def clean_list_record(self, record):
if record['county'].upper() not in self.county_names:
raise SkipRecord('Skipping county %s' % record['county'])
record['approval_date'] = parse_date(record['approval_date'], self.date_format)
if record['approval_date'] is None:
raise SkipRecord('Record has no date.')
return record
def clean_list_record(self, record):
notes = []
notes_pats = [r'(?P<value>.*?)\s*\-*\s*(?P<notes>\(?\s*w\/d.*)',
r'(?P<value>.*?)\s*\-*\s*(?P<notes>\(?\s*withd.*)',
r'(?P<value>.*?)\s*\-*\s*(?P<notes>\(?\s*ch\s+227\s+sec\s*5A.*)',
r'(?P<value>.*?)\s*\-*\s*(?P<notes>\(?\s*ch\s+bus\s+.*)',
r'(?P<value>.*?)\s*\-*\s*(?P<notes>\(?\s*c\/l.*)',
]
# strip notes off of several cruft-prone fields
for field in ['name', 'business_type', 'location']:
val = record.get(field, '').strip()
for pat in notes_pats:
m = re.match(pat, val, re.I|re.M)
if m is not None:
results = m.groupdict()
val = results['value']
notes.append(results['notes'])
record[field] = val.strip()
record['notes'] = notes
record['location'] = smart_title(record['location'].strip())
record['date'] = parse_date(record['date'].strip(), '%Y-%m-%d')
if (record['name'].upper(), record['location'].upper()) in BUSINESS_NAMES_TO_IGNORE:
raise SkipRecord('Skipping %s (explicitly ignored)' % record['name'])
if (record['location'] == ''):
raise SkipRecord('Skipping %s (no location)' % record['name'])
return record
def clean_list_record(self, record):
# Get the first 2 digits of the UCR, or None.
ucr1_prefix = record['offenseucr1'] and record[
'offenseucr1'][:2] or None
ucr2_prefix = record['offenseucr2'] and record[
'offenseucr2'][:2] or None
if ucr1_prefix in SKIPPED_UCR_PREFIXES:
raise SkipRecord('Skipping record with UCR %s' %
record['offenseucr1'])
if ucr2_prefix in SKIPPED_UCR_PREFIXES:
raise SkipRecord('Skipping record with UCR %s' %
record['offenseucr2'])
record['address'] = self.normalizer.normalize_address(record)
# If we can't find the code in the mapping, just use the code itself
# as the value.
record['category'] = CATEGORY_MAPPING.get(ucr1_prefix, ucr1_prefix)
record['crime_type'] = UCR_MAPPING.get(record['offenseucr1'],
record['offenseucr1'])
record['secondary_category'] = CATEGORY_MAPPING.get(
ucr2_prefix, ucr2_prefix)
record['secondary_crime_type'] = UCR_MAPPING.get(
record['offenseucr2'], record['offenseucr2'])
record['offensedate'] = parse_date(record['offensedate'], '%m/%d/%Y')
record['offensestarttime'] = datetime.time(
*time.strptime(record['offensestarttime'], '%H:%M:%S')[3:5])
record['offensebeat'] = record['offensebeat'] or ''
return record
def clean_list_record(self, record):
if record['WORKDESC'] == '4635':
raise SkipRecord('WORKDESC is 4635')
if record['JOBLOCATION'] is None:
raise SkipRecord('Record has no location')
if record['PERMITAPPROVALS'] is None:
record['permit_types'] = []
else:
record['permit_types'] = [
self.clean_permit_type(pa)
for pa in record['PERMITAPPROVALS'].split(',')
]
# Addresses and extra data generally seem to be separated by 2 or more spaces
record['location'] = record['JOBLOCATION'].split(' ')[0]
for format in ['%d-%b-%y', '%m/%d/%Y']:
try:
issue_date = parse_date(record['ISSUEDATE'], format)
break
except ValueError:
continue
record['issue_date'] = issue_date
record['WORKDESC'] = record['WORKDESC'].strip()
record['SUBDESC'] = record['SUBDESC'].strip()
return record
def clean_list_record(self, record):
record['address'] = '%s %s, %s' % (record['House #'],
smart_title(record['Street Name']),
smart_title(record['Borough']))
record['is_landmark'] = record['Landmark'] == 'Y'
record['is_adult_establishment'] = record['Adult Estab'] == 'Y'
record['is_city_owned'] = record['City Owned'] == 'Y'
record['illumination_type'] = record.get(
'Sign Illumination Type', 'Not available') or 'Not illuminated'
if not isinstance(record['Latest Action Date'], datetime.datetime):
self.logger.info('Skipping job #%s, with latest action date %s',
record.get('Job #'), record['Latest Action Date'])
raise SkipRecord()
try:
record['sign_text'] = record['Text on Sign'].strip()
if len(record['sign_text']) > 255:
# Some records are malformed and have a bad and long value
# for sign text.
self.logger.info('Skipping job #%s, with Text on Sign %s',
record.get('Job #'), record['Text on Sign'])
raise SkipRecord()
except AttributeError:
try:
record['sign_text'] = str(int(record['Text on Sign']))
except TypeError:
self.logger.info('Skipping job #%s, with Text on Sign %s',
record.get('Job #'), record['Text on Sign'])
raise SkipRecord()
try:
record['sign_for'] = record['Sign Advertising']
except KeyError:
record['sign_for'] = record['Usage']
try:
record['is_near_highway'] = record['Sign Near Highway'] == 'Y'
except KeyError:
# Older spreadsheets don't have a 'Sign Near Highway' column,
# and there's nothing we can do about it. They have a column called
# 'Adjacent to Arterial Highway', but that's not necessarily the
# same thing.
record['is_near_highway'] = None
try:
record['is_changeable_copy'] = record[
'Sign Changeable Copy'] == 'Y'
except KeyError:
# Older spreadsheets don't have a 'Sign Changeable Copy' column,
# but we can deduce the value: if there's text, then it's not
# changeable. Otherwise, it's NULL.
if record['sign_text']:
record['is_changeable_copy'] = False
else:
record['is_changeable_copy'] = None
return record
def clean_list_record(self, rss_record):
record = {
'pub_date': datetime.date(*rss_record.pop('updated_parsed')[:3]),
'summary': rss_record['summary'].strip(),
}
if re.search(r'^(?i)\*UPDATE:', record['summary']):
m = re.search(
r'^\*UPDATE:\s*(?P<location_name>[^\*]*)\*\s*(?P<description>.*)\s*-\s*(?P<reporter>.*?)\#\#\#$',
record['summary'])
if not m:
self.logger.warn('Could not parse update %r' %
record['summary'])
raise SkipRecord('Could not parse update %r' %
record['summary'])
record.update(m.groupdict())
record.update({
'is_update': True,
'incident_type': '',
'fire_station': '',
'radio_channels': '',
'incident_time': '',
})
else: # Not an update
m = re.search(
r'^\*(?P<incident_type>[^\*]*)\*\s*(?P<location_name>[^;]*);\s*MAP (?:\d+[- ]\w\d)?;\s*FS (?P<fire_station>\d+); (?P<description>.*?); Ch:(?P<radio_channels>[\d, ]+)\s*@(?P<incident_time>\d\d?:\d\d [AP]M)?\s*-(?P<reporter>.*?)\#\#\#$',
record['summary'])
if not m:
raise SkipRecord('Could not parse %r' % record['summary'])
record.update(m.groupdict())
record['incident_type'] = record['incident_type'].upper(
) # Normalize
record['radio_channels'] = ','.join(
record['radio_channels'].split(','))
record['is_update'] = False
record['description'] = record['description'].replace(
' ', ' ').replace('"', '"').replace('&',
'&').strip()
record['location_name'] = record['location_name'].strip()
# Get the incident ID and message ID from the Google Groups URL.
# We'll use these as unique identifiers.
m = re.search(
r'browse_thread/thread/(?P<incident_id>[^/]*)/(?P<message_id>[^\?]*)\?',
rss_record['link'])
if not m:
raise ScraperBroken('Got weird URL: %r', rss_record['link'])
record.update(m.groupdict())
record['link'] = rss_record['link']
# I can't figure out why this record is causing errors, so for now
# we'll just skip it.
if record['message_id'] == '0faabeab3aad8492':
raise SkipRecord()
return record
def clean_list_record(self, record):
if record['lat'] == '0' or record['long'] == '0':
raise SkipRecord('Got 0 for lat/long')
record['lat'] = float(record['lat'])
record['long'] = float(record['long'])
record['location_name'] = '%s %s, from %s to %s' % (
record['street'], record['quadrant'], record['fromintersection'],
record['tointersection'])
# record['order_date'] = parse_date(record['serviceorderdate'].split('T')[0], '%Y-%m-%d')
# record['add_date'] = parse_date(record['adddate'].split('T')[0], '%Y-%m-%d')
if record['estimatedcompletiondate']:
record['estimated_completion_date'] = parse_date(
record['estimatedcompletiondate'].split('T')[0], '%Y-%m-%d')
else:
record['estimated_completion_date'] = None
if record['actualcompletiondate']:
record['actual_completion_date'] = parse_date(
record['actualcompletiondate'].split('T')[0], '%Y-%m-%d')
else:
record['actual_completion_date'] = None
if record['estimatedstartdate']:
record['estimated_start_date'] = parse_date(
record['estimatedstartdate'].split('T')[0], '%Y-%m-%d')
else:
record['estimated_start_date'] = None
record['project_name'] = record['projectname'] or ''
record['road_type'] = record['functionalclass'] or 'N/A'
return record
def clean_list_record(self, record):
# clean up a record dict
venue = record.get('venue', {})
if not venue:
raise SkipRecord("No venue")
location_name_parts = [
venue.get(key, '').strip()
for key in ('address_1', 'address_2', 'city', 'state', 'zip')
]
location_name = ', '.join([p for p in location_name_parts if p])
event_time = datetime.datetime.fromtimestamp(record['time'] / 1000.0,
local_tz)
cleaned = {
'title': text_from_html(record['name']),
'description': text_from_html(record.get('description', '')),
'location_name': location_name,
'location': Point(venue['lon'], venue['lat']),
'url': record['event_url'],
'item_date': event_time.date(),
}
attributes = {
'venue_phone': venue.get('phone', ''),
'venue_name': text_from_html(venue.get('name', '')),
'start_time': event_time.time(),
'group_name': record['group']['name'],
}
cleaned['_attributes'] = attributes
return cleaned
def clean_list_record(self, record):
if record['county'].upper().strip() not in self.counties:
raise SkipRecord('Record not in %s.' % self.counties)
record['activity_date'] = parse_date(record['activity_date'],
'%m/%d/%Y')
record['dba'] = smart_title(record['dba'])
record['address'] = clean_address(record['address'])
return record
def clean_list_record(self, record):
record = super(Scraper, self).clean_list_record(record)
# Special-case: a bad record identified on 2009-02-11.
if str(record['inspection_visit_id']) == '3145783' and str(record['license_id']) == '2164597':
raise SkipRecord('Invalid record')
return record
def clean_list_record(self, record):
record['location'] = self.get_location(record['PERMIT_NUMBER'])
if record['location'] is None:
raise SkipRecord("No permit found for '%s'" %
record['PERMIT_NUMBER'])
record['InspectionDate'] = parse_date(record['InspectionDate'],
'%Y-%m-%dT00:00:00')
return record
def clean_list_record(self, record):
record['name'] = record['name'].strip()
record['business_type'] = record['business_type'].strip()
record['location'] = smart_title(record['location'].strip())
record['date'] = parse_date(record['date'].strip(), '%Y-%m-%d')
if (record['name'].upper(),
record['location'].upper()) in BUSINESS_NAMES_TO_IGNORE:
raise SkipRecord('Skipping %s' % record['name'])
return record
def clean_list_record(self, record):
if record['lat'] == '0' or record['long'] == '0':
raise SkipRecord('Got 0 for lat/long')
record['lat'] = float(record['lat'])
record['long'] = float(record['long'])
record['report_date'] = parse_date(record['reportdatetime'].split('T')[0], '%Y-%m-%d')
record['address'] = record['blocksiteaddress'].replace(' B/O ', ' block of ')
if record['narrative'].upper() == 'NO NARRATIVE IS AVAILABLE.':
record['narrative'] == 'Not available'
return record
def normalize_address(self, record):
"""
Addresses are provided with no spaces, so try to find the suffix if
there is one, then compare the reamining part to the streets table.
"""
if record['offensestreet'] is None:
raise SkipRecord('Skipping record with no street')
street = record['offensestreet']
matching_suffix = ''
suffix_groups = [('EXPWY', 'PKWY', 'FRWY', 'BLVD'),
('HWY', 'FRW', 'AVE', 'CIR', 'EXT', 'BLV', 'PKW',
'ROW', 'WAY', 'EXP'),
('DR', 'ST', 'RD', 'LN', 'BL', 'TR', 'WY', 'CT', 'PL',
'AV', 'CI'), ('P', 'R', 'F', 'D', 'S', 'L')]
match_found = False
for group in suffix_groups:
if match_found:
break
for suffix in group:
if record['offensestreet'].endswith(suffix):
street_name = record['offensestreet'][:-len(suffix)]
# Try looking up the street name from a dictionary mapping
# collapsed street names to names in the streets table.
try:
street = self.streets[street_name]
matching_suffix = suffix
match_found = True
break
except KeyError:
# SAINT is encoded as ST in the data, but Saint in the streets table,
# so try that if the address starts with ST.
if street_name.startswith('ST'):
street_name = 'SAINT%s' % street_name[2:]
try:
street = self.streets[street_name]
matching_suffix = suffix
match_found = True
break
except KeyError:
continue
if match_found:
self.matches_found += 1
self.records_seen += 1
normalized_block = record['offenseblock'].lstrip('0')
if normalized_block[-2:] == 'xx':
normalized_block = normalized_block.replace('xx', '00')
address = '%s %s %s %s' % (normalized_block, record['offensedirection']
or '', street, matching_suffix)
address = re.sub(r'\s+', ' ', address)
return address_to_block(address)
def clean_list_record(self, record):
record['address'] = '%s %s, %s' % (record['House #'],
smart_title(record['Street Name']),
smart_title(record['Borough']))
record['is_landmark'] = record['Landmarked'] == 'Y'
record['is_adult_establishment'] = record['Adult Estab'] == 'Y'
record['is_city_owned'] = record['City Owned'] == 'Y'
if not isinstance(record['Latest Action Date'],
(datetime.date, datetime.datetime)):
raise SkipRecord('Got last action date %s' %
record['Latest Action Date'])
return record
def clean_list_record(self, record):
try:
dt = strptime(record['date'].strip(), '%m/%d/%Y %I:%M:%S %p')
record['incident_date'] = datetime.date(*dt[:3])
record['incident_time'] = datetime.time(*dt[3:6])
except ValueError:
dt = strptime(record['date'].strip(), '%m/%d/%Y')
record['incident_date'] = datetime.date(*dt[:3])
record['incident_time'] = None
record['units'] = record['units'].split()
if 'mutual aid' in record['type'].lower():
raise SkipRecord('Skipping mutual aid: %r' % record['type'])
return record
def clean_list_record(self, record):
for k, v in record.items():
v = strip_tags(v)
record[k] = re.sub(r'(?s)\s\s+', ' ', v).strip()
# Remove the "Suite/Apt" or "Floor" clause from the address, if it exists.
record['address'] = record['address'].replace(' ,', ',')
m = re.search(r'^(.*?), (?:Suite/Apt|Floor):.*$', record['address'])
if m:
record['address'] = m.group(1)
record['address'] = clean_address(record['address'])
if record['dba'] in SKIPPED_DBAS:
raise SkipRecord('Skipping %r' % record['dba'])
# For privacy reasons, skip individuals.
if record['structure'].upper().strip() == 'INDIVIDUAL':
raise SkipRecord('Skipping structure=individual')
record['city_id'] = int(record['city_id'])
record['site_id'] = int(record['site_id'])
return record
def clean_list_record(self, record):
record['application_date'] = parse_date(record['application_date'], '%m/%d/%Y')
try:
record['review_date'] = parse_date(record['review_date'], '%m/%d/%Y')
except ValueError: # sometimes it's 'n/a'
record['review_date'] = None
record['address'] = strip_unit(clean_address(record['address']))
if record['city'] not in self.city_names:
raise SkipRecord('Skipping city %s' % record['city'])
record['city'] = record['city'].title()
return record
def parse_detail(self, page, list_record):
detail_record = []
for m in self.parse_detail_re.finditer(page):
record = m.groupdict()
if record['inspection_type'] in INSPECTION_TYPES_TO_IGNORE:
raise SkipRecord('Got ignorable inspection type %r' %
record['inspection_type'])
html = record.pop('html')
record['violations'] = []
for vm in self.parse_violations_re.finditer(html):
record['violations'].append(vm.group(1))
detail_record.append(record)
return detail_record
def clean_list_record(self, record):
# The facility type is determied by the 6th and 7th digits in the
# facility ID.
facility_type = record['FAC_ID'][5:7]
if facility_type in EXCLUDED_FACILITY_TYPES:
raise SkipRecord('Excluding record from facility type %s' %
facility_type)
record['DATE'] = parse_date(record['DATE']).date()
record['FAC_NAME'] = record['FAC_NAME'].decode('Latin-1')
record['FIN_SCORE'] = float(record['FIN_SCORE']) + float(
record['RAW_SCORE'])
record['schema_slug'] = self.get_schema_slug(record)
record['facility_type'] = self.clean_facility_type(facility_type)
record['result'] = self.get_result(record)
return record
def clean_list_record(self, record):
record.title = convert_entities(record['title'])
record.description = convert_entities(record['description'])
# Don't know why, but some feeds have 'id' *instead* of 'link'.
if record.get('id', '').startswith('http'):
record['link'] = record['id']
# This tries GeoRSS, RDF Geo, xCal, ...
point, location_name = self.get_point_and_location_name(record)
_short_title = record['title'][:30] + '...'
if not point:
raise SkipRecord("couldn't geocode any addresses in item '%s...'" %
_short_title)
if not location_name:
raise SkipRecord(
"Skip, no location name and failed to reverse geocode %s for %r"
% (point.wkt, _short_title))
if not intersects_metro_bbox(point):
# Check if latitude, longitude seem to be reversed; I've
# seen that in some bad feeds!
reversed_loc = Point(point.y, point.x)
if intersects_metro_bbox(reversed_loc):
self.logger.info(
"Got points in apparently reverse order, flipping them")
point = reversed_loc
else:
raise SkipRecord("Skipping %r as %s,%s is out of bounds" %
(_short_title, point.y, point.x))
record['location_name'] = location_name
record['location'] = point
return record
def save(self, old_record, list_record, detail_record):
attributes = list_record.pop('attributes', {})
list_record.setdefault('schema', self.schema.id)
if not old_record:
list_record.setdefault('item_date', today())
list_record.setdefault('pub_date', now())
from ebpub.db.forms import NewsItemForm
form = NewsItemForm(list_record, instance=old_record)
if form.is_valid():
return self.create_or_update(old_record, attributes,
**form.cleaned_data)
else:
self.logger.info("Skipping due to validation failures:")
for key, val in form.errors.items():
self.logger.info("%s: %s" % (key, val.as_text()))
raise SkipRecord(form.errors)
def clean_list_record(self, record):
record['waiver_date'] = parse_date(record['waiver_date'], '%m/%d/%y')
record['address'] = (
'%s %s %s' % (record.pop('street_number', '').strip(),
record.pop('street_name', '').strip(),
record.pop('street_suffix', '').strip())).strip()
try:
record['borough'] = {
'BK': 'Brooklyn',
'BX': 'The Bronx',
'MN': 'Manhattan',
'QS': 'Queens',
'SI': 'Staten Island',
}[record['borough']]
except KeyError:
raise SkipRecord('Invalid borough')
return record
def clean_list_record(self, record):
for k, v in record.items():
record[k] = strip_unneeded_tags(v)
if not record['street_name']:
raise SkipRecord('Street name not found')
record['start_date'] = parse_date(record['start_date'], '%m/%d/%y')
record['end_date'] = parse_date(record['end_date'], '%m/%d/%y')
record['date_posted'] = parse_date(
'%s_%s' % (record.pop('update_date'), record.pop('update_time')),
'%m/%d/%Y_%I:%M %p',
return_datetime=True)
record['address'] = '%s-%s %s %s %s' % (
record['block_from'], record['block_to'], record['street_dir'],
record['street_name'], record['street_suffix'])
record['details'] = re.sub(r'(?i)<br>', '\n',
record['details']).replace(' ',
' ').strip()
return record
def clean_list_record(self, record):
# clean up a record dict
# Item date, in timezone of the photo owner.
# Not sure how to determine what that is, so we'll leave it.
cleaned = {}
cleaned['item_date'] = datetime.datetime.strptime(
record['datetaken'], '%Y-%m-%d %H:%M:%S')
cleaned['item_date'] = cleaned['item_date'].date()
# Posted date, UTC timestamp.
pub_date = datetime.datetime.fromtimestamp(float(record['dateupload']),
utc)
cleaned['pub_date'] = pub_date.astimezone(local_tz)
description = record['description']['_content']
cleaned['description'] = convert_entities(description.strip())
cleaned['title'] = convert_entities(record['title'])
x, y = record['longitude'], record['latitude']
cleaned['location'] = Point((float(x), float(y)))
# Possibly we could figure out flickr's geo API and resolve
# the photo's place_id and/or woeid to the place name? But
# those are probably not specific enough; reverse-geocode
# instead.
try:
block, distance = reverse_geocode(cleaned['location'])
cleaned['location_name'] = block.pretty_name
except ReverseGeocodeError:
raise SkipRecord("Could not geocode location %s, %s" % (x, y))
# Don't think any of the urls returned by the API's "extras"
# correspond to the page? not sure.
cleaned[
'url'] = 'http://www.flickr.com/photos/%(owner)s/%(id)s' % record
attributes = {}
attributes['sourcename'] = 'Flickr'
#attributes['photo_id'] = record['id']
attributes['user_id'] = record['owner']
attributes['username'] = record['ownername']
# Thumbnail. 'Small square' photos are 75x75.
attributes['photo_href'] = record['url_sq']
cleaned['_attributes'] = attributes
return cleaned
def clean_list_record(self, record):
if record['last_inspection_date'].lower() == 'not available':
raise SkipRecord('No inspection available')
else:
record['last_inspection_date'] = parse_date(
record['last_inspection_date'], '%m/%d/%Y')
if record['aka']:
record['aka'] = list_aka_re.findall(record['aka'])[0]
else:
record['aka'] = ''
norm_dict_space(record, 'name', 'dba', 'address')
record['result'] = record['result'].replace(' ', '').strip()
record['city_id'] = int(record['city_id'])
# Remove the trailing ZIP code from the address, if it exists.
m = re.search(r'(.*?)\s+\d\d\d\d\d$', record['address'])
if m:
record['address'] = m.group(1)
record['address'] = clean_address(record['address'])
return record
def clean_list_record(self, record):
record['DateOfInspection'] = parse_date(record['DateOfInspection'], '%Y-%m-%d')
if not record['DateOfInspection']:
raise SkipRecord('Inspection date not given')
record['Total Time'] = record['Total Time'].strip() or None
record['EmployeeIDofInspector'] = record['EmployeeIDofInspector'].strip() or None
# Calculate the score range.
record['Score'] = record['Score'].strip() or None
if record['Score']:
try:
record['Score'] = int(record['Score'])
except ValueError:
raise SkipRecord('Got sketchy non-integer score %r' % record['Score'])
if record['Score'] >= 91:
record['score_range'] = '91-100'
elif record['Score'] >= 81:
record['score_range'] = '81-90'
elif record['Score'] >= 71:
record['score_range'] = '71-80'
elif record['Score'] >= 61:
record['score_range'] = '61-70'
elif record['Score'] >= 51:
record['score_range'] = '51-60'
elif record['Score'] >= 41:
record['score_range'] = '41-50'
elif record['Score'] >= 31:
record['score_range'] = '31-40'
elif record['Score'] >= 21:
record['score_range'] = '21-30'
elif record['Score'] >= 11:
record['score_range'] = '11-20'
else:
record['score_range'] = '0-10'
else:
record['score_range'] = 'N/A'
# Get the location data from the locations table.
try:
loc = self.locations()[record['LocationIDInspected']]
except KeyError:
raise SkipRecord('Location not found')
# Explicitly convert to a Unicode object using the utf8 codec, because
# this might contain funky characters.
record['restaurant_dba'] = loc['DBA'].decode('utf8').strip()
record['restaurant_type'] = loc['Type Description'].strip()
record['address'] = loc['StreetAddress'].decode('utf8').strip()
record['LocationIDInspected'] = int(record['LocationIDInspected'])
# Get the inspection type data from the inspection types table.
try:
inspection_type = self.inspection_types()[record['InspectionTypeID']]
except KeyError:
raise SkipRecord('Inspection type not found')
record['inspection_type'] = inspection_type['InspectionType']
# Get the violation data from the violations table.
try:
vios = self.violations()[record['InspectionID']]
except KeyError:
record['violations'] = []
else:
vio_types = self.violation_types()
record['violations'] = [dict(vio, type=vio_types[vio['ViolationTypeID']]) for vio in vios]
return record
def clean_list_record(self, record):
if record['title'].startswith(u'Boston 24'):
# We don't include the summary posts.
# TODO: the 'Boston 24' tag indicates posts with aggregate
# daily stats. Make a separate schema for the aggregates,
# with attributes like those used in
# everyblock/everyblock/cities/nyc/crime_aggregate/retrieval.py.
# Or maybe not: these are citywide, not by precinct.
# So what would be the Location? Whole city??
self.logger.info("boston daily crime stats, we don't know how to "
"handle these yet")
raise SkipRecord
date = datetime.date(*record['updated_parsed'][:3])
description = record['summary']
# This feed doesn't provide geographic data; we'll try to
# extract addresses from the text, and stop on the first
# one that successfully geocodes.
full_description = record['content'][0]['value']
full_description = text_from_html(full_description)
location, location_name = self.get_point_and_location_name(
record, address_text=full_description)
if not (location or location_name):
raise SkipRecord("No location or location_name")
# Get the precinct from the tags.
precincts = [
'A1', 'A15', 'A7', 'B2', 'B3', 'C11', 'C6', 'D14', 'D4', 'E13',
'E18', 'E5'
]
tags = [t['term'] for t in record['tags']]
precinct = None
for tag in tags:
if tag in precincts:
# TODO: we need a LocationType for precincts, and shapes; and
# then we could set newsitem.location_object to the Location
# for this precinct. For now we just save it as an attribute.
precinct = tag
break
attributes = {}
if precinct:
precinct = self.get_or_create_lookup('precinct', precinct,
precinct)
attributes['precinct'] = precinct.id
else:
raise SkipRecord("no precinct found in tags %r" % tags)
cleaned = dict(
item_date=date,
location=location,
location_name=location_name,
title=record['title'],
description=description,
url=record['link'],
attributes=attributes,
)
return cleaned
def clean_list_record(self, record):
if record['primary_status'] == 'Acknowledged':
raise SkipRecord('Status is "Acknowledged"')
return super(ConvertedCondominiumsScraper, self).clean_list_record(record)
