def clean_list_record(self, record):
record['CALL_DATE'] = parse_date(record['CALL_DATE'], '%Y-%m-%d')
record['OFFENSE_DATE'] = parse_date(record['OFFENSE_DATE'], '%Y-%m-%d')
record['REPORT_DATE'] = parse_date(record['REPORT_DATE'], '%Y-%m-%d')
record['COMMON_LOCATION'] = record['COMMON_LOCATION'].strip()
record['ADDRESS_NBR'] = record['ADDRESS_NBR'].strip()
record['ADDRESS'] = record['ADDRESS'].strip()
# The 'NARRATIVE' field includes time and disposition data. Parse that out.
m = re.search(
r'^Time: (?P<TIME>\d\d?:\d\d)<br>.*?<br>Disposition: (?P<DISPOSITION>.*)$',
record.pop('NARRATIVE'))
record.update(m.groupdict())
record['TIME'] = parse_time(record['TIME'], '%H:%M')
# Set location_name. The logic is different depending on the ADDRESS_TYPE.
address_type = record['ADDRESS_TYPE']
if address_type == 'PREMISE ADDRESS':
record['location_name'] = '%s block of %s' % (
record['ADDRESS_NBR'], clean_address(record['ADDRESS']))
elif address_type == 'INTERSECTION':
if '/' in record['ADDRESS']:
streets = record['ADDRESS'].split('/')
record['location_name'] = '%s and %s' % (clean_address(
streets[0]), clean_address(streets[1]))
else:
record['location_name'] = clean_address(record['ADDRESS'])
elif address_type == 'GEO-OVERRIDE':
record['location_name'] = clean_address(record['ADDRESS'])
elif address_type == 'COMMON LOCATION':
if record['ADDRESS_NBR'] and record['ADDRESS']:
record['location_name'] = '%s %s' % (
record['ADDRESS_NBR'], clean_address(record['ADDRESS']))
elif record['ADDRESS'] and record['COMMON_LOCATION']:
record['location_name'] = '%s (%s)' % (clean_address(
record['ADDRESS']), clean_address(
record['COMMON_LOCATION']))
elif record['COMMON_LOCATION']:
record['location_name'] = clean_address(
record['COMMON_LOCATION'])
elif record['ADDRESS']:
record['location_name'] = clean_address(record['ADDRESS'])
else:
record['location_name'] = 'Unknown'
else:
record['location_name'] = 'Unknown'
try:
d = CUSTOM_CATEGORIES[record['ORIG_CRIMETYPE_NAME']]
except KeyError:
d = ('Unknown', 'Unknown')
record['broad_category'], record['detail_category'] = d
return record
def clean_list_record(self, record):
record['CALL_DATE'] = parse_date(record['CALL_DATE'], '%Y-%m-%d')
record['OFFENSE_DATE'] = parse_date(record['OFFENSE_DATE'], '%Y-%m-%d')
record['REPORT_DATE'] = parse_date(record['REPORT_DATE'], '%Y-%m-%d')
record['COMMON_LOCATION'] = record['COMMON_LOCATION'].strip()
record['ADDRESS_NBR'] = record['ADDRESS_NBR'].strip()
record['ADDRESS'] = record['ADDRESS'].strip()
# The 'NARRATIVE' field includes time and disposition data. Parse that out.
m = re.search(r'^Time: (?P<TIME>\d\d?:\d\d)<br>.*?<br>Disposition: (?P<DISPOSITION>.*)$', record.pop('NARRATIVE'))
record.update(m.groupdict())
record['TIME'] = parse_time(record['TIME'], '%H:%M')
# Set location_name. The logic is different depending on the ADDRESS_TYPE.
address_type = record['ADDRESS_TYPE']
if address_type == 'PREMISE ADDRESS':
record['location_name'] = '%s block of %s' % (record['ADDRESS_NBR'], clean_address(record['ADDRESS']))
elif address_type == 'INTERSECTION':
if '/' in record['ADDRESS']:
streets = record['ADDRESS'].split('/')
record['location_name'] = '%s and %s' % (clean_address(streets[0]), clean_address(streets[1]))
else:
record['location_name'] = clean_address(record['ADDRESS'])
elif address_type == 'GEO-OVERRIDE':
record['location_name'] = clean_address(record['ADDRESS'])
elif address_type == 'COMMON LOCATION':
if record['ADDRESS_NBR'] and record['ADDRESS']:
record['location_name'] = '%s %s' % (record['ADDRESS_NBR'], clean_address(record['ADDRESS']))
elif record['ADDRESS'] and record['COMMON_LOCATION']:
record['location_name'] = '%s (%s)' % (clean_address(record['ADDRESS']), clean_address(record['COMMON_LOCATION']))
elif record['COMMON_LOCATION']:
record['location_name'] = clean_address(record['COMMON_LOCATION'])
elif record['ADDRESS']:
record['location_name'] = clean_address(record['ADDRESS'])
else:
record['location_name'] = 'Unknown'
else:
record['location_name'] = 'Unknown'
try:
d = CUSTOM_CATEGORIES[record['ORIG_CRIMETYPE_NAME']]
except KeyError:
d = ('Unknown', 'Unknown')
record['broad_category'], record['detail_category'] = d
return record
def clean_list_record(self, record):
record['crime_date'] = parse_date(record['crime_date'], '%m-%d-%Y')
record['crime_time'] = parse_time(record['crime_time'], '%I:%M:%S %p')
record['lon'] = float(record['lon'])
record['lat'] = float(record['lat'])
return record
def clean_list_record(self, record):
record['crime_date'] = parse_date(record['crime_date'], '%m-%d-%Y')
record['crime_time'] = parse_time(record['crime_time'], '%I:%M:%S %p')
record['lon'] = float(record['lon'])
record['lat'] = float(record['lat'])
return record
def _item_create(info):
info = copy.deepcopy(info)
try:
assert info.pop('type') == 'Feature'
props = info['properties']
except (KeyError, AssertionError):
raise InvalidNewsItem({'type': 'not a valid GeoJSON Feature'})
try:
slug = props.pop('type', None)
schema = models.Schema.objects.get(slug=slug)
except (models.Schema.DoesNotExist):
raise InvalidNewsItem({'type': 'schema %r does not exist' % slug})
data = {'schema': schema.id}
for key in ('title', 'description', 'url'):
data[key] = props.pop(key, '')
# If there are errors parsing the dates, keep the raw data and let
# the ModelForm sort it out.
pub_date = props.pop('pub_date', None)
if pub_date:
try:
data['pub_date'] = normalize_datetime(pyrfc3339.parse(pub_date))
except Exception:
data['pub_date'] = pub_date
else:
data['pub_date'] = normalize_datetime(datetime.datetime.utcnow())
item_date = props.pop('item_date', None)
if item_date:
try:
data['item_date'] = parse_date(item_date, '%Y-%m-%d', False)
except Exception:
data['item_date'] = item_date
else:
try:
data['item_date'] = data['pub_date'].date()
except Exception:
data['item_date'] = None
data['location'], data['location_name'] = _get_location_info(
info.get('geometry'), props.pop('location_name', None))
if not data['location']:
logger.warn("Saving NewsItem %s with no geometry" % data['title'])
if not data['location_name']:
logger.warn("Saving NewsItem %s with no location_name" % data['title'])
from ebpub.db.forms import NewsItemForm
form = NewsItemForm(data)
if form.is_valid():
item = form.save()
else:
raise InvalidNewsItem(form.errors)
# Everything else goes in .attributes.
attributes = {}
for key, val in props.items():
sf = models.SchemaField.objects.get(schema=schema, name=key)
if sf.is_many_to_many_lookup():
lookups = []
for lookup_name in val:
lookups.append(
models.Lookup.objects.get_or_create_lookup(sf, lookup_name))
val = ','.join((str(lookup.id) for lookup in lookups))
elif sf.is_lookup:
val = models.Lookup.objects.get_or_create_lookup(sf, val)
elif sf.is_type('date'):
val = normalize_datetime(parse_date(val, '%Y-%m-%d'))
elif sf.is_type('time'):
val = normalize_datetime(parse_time(val, '%H:%M'))
elif sf.is_type('datetime'):
val = normalize_datetime(pyrfc3339.parse(datetime))
attributes[key] = val
item.attributes = attributes
return item
def _item_create(info):
info = copy.deepcopy(info)
try:
assert info.pop('type') == 'Feature'
props = info['properties']
except (KeyError, AssertionError):
raise InvalidNewsItem({'type': 'not a valid GeoJSON Feature'})
try:
slug = props.pop('type', None)
schema = models.Schema.objects.get(slug=slug)
except (models.Schema.DoesNotExist):
raise InvalidNewsItem({'type': 'schema %r does not exist' % slug})
data = {'schema': schema.id}
for key in ('title', 'description', 'url'):
data[key] = props.pop(key, '')
# If there are errors parsing the dates, keep the raw data and let
# the ModelForm sort it out.
pub_date = props.pop('pub_date', None)
if pub_date:
try:
data['pub_date'] = normalize_datetime(pyrfc3339.parse(pub_date))
except Exception:
data['pub_date'] = pub_date
else:
data['pub_date'] = normalize_datetime(datetime.datetime.utcnow())
item_date = props.pop('item_date', None)
if item_date:
try:
data['item_date'] = parse_date(item_date, '%Y-%m-%d', False)
except Exception:
data['item_date'] = item_date
else:
try:
data['item_date'] = data['pub_date'].date()
except Exception:
data['item_date'] = None
data['location'], data['location_name'] = _get_location_info(
info.get('geometry'), props.pop('location_name', None))
if not data['location']:
logger.warn("Saving NewsItem %s with no geometry" % data['title'])
if not data['location_name']:
logger.warn("Saving NewsItem %s with no location_name" % data['title'])
from ebpub.db.forms import NewsItemForm
form = NewsItemForm(data)
if form.is_valid():
item = form.save()
else:
raise InvalidNewsItem(form.errors)
# Everything else goes in .attributes.
attributes = {}
for key, val in props.items():
sf = models.SchemaField.objects.get(schema=schema, name=key)
if sf.is_many_to_many_lookup():
lookups = []
for lookup_name in val:
lookups.append(
models.Lookup.objects.get_or_create_lookup(
sf, lookup_name))
val = ','.join((str(lookup.id) for lookup in lookups))
elif sf.is_lookup:
val = models.Lookup.objects.get_or_create_lookup(sf, val)
elif sf.is_type('date'):
val = normalize_datetime(parse_date(val, '%Y-%m-%d'))
elif sf.is_type('time'):
val = normalize_datetime(parse_time(val, '%H:%M'))
elif sf.is_type('datetime'):
val = normalize_datetime(pyrfc3339.parse(datetime))
attributes[key] = val
item.attributes = attributes
return item
def clean_list_record(self, record):
record["crime_date"] = parse_date(record["crime_date"], "%m-%d-%Y")
record["crime_time"] = parse_time(record["crime_time"], "%I:%M:%S %p")
record["lon"] = float(record["lon"])
record["lat"] = float(record["lat"])
return record
