def provider_map(delete=False):
''' Adds favicons and metadata for harvesters to Elasticsearch '''
from six.moves.urllib import parse as urllib_parse
from scrapi import registry
from scrapi.base.helpers import null_on_error
from scrapi.processing.elasticsearch import DatabaseManager
dm = DatabaseManager()
dm.setup()
es = dm.es
if delete:
es.indices.delete(index='share_providers', ignore=[404])
from scrapi.harvesters.push_api import gen_harvesters
gen_harvesters()
for harvester_name, harvester in registry.items():
if not null_on_error(es.get, log=False)(index='share_providers', doc_type=harvester_name, id=harvester_name):
with open("img/favicons/{}_favicon.ico".format(harvester.short_name), "rb") as f:
favicon = urllib_parse.quote(base64.encodestring(f.read()))
es.index(
'share_providers',
harvester.short_name,
body={
'favicon': 'data:image/png;base64,' + favicon,
'short_name': harvester.short_name,
'long_name': harvester.long_name,
'url': harvester.url
},
id=harvester.short_name,
refresh=True
)
print(es.count('share_providers', body={'query': {'match_all': {}}})['count'])
def format_property(self, property):
if property == "date":
fn = compose(
lambda x: list(map(null_on_error(datetime_formatter), x)), coerce_to_list, self.resolve_property
)
else:
fn = self.resolve_property
return (property, ("//dc:{}/node()".format(property), "//ns0:{}/node()".format(property), fn))
def format_property(self, property):
if property == 'date':
fn = compose(
lambda x: list(map(null_on_error(datetime_formatter), x)),
coerce_to_list, self.resolve_property)
else:
fn = self.resolve_property
return (property, ('//dc:{}/node()'.format(property),
'//ns0:{}/node()'.format(property), fn))
def format_property(self, property):
if property == 'date':
fn = compose(lambda x: map(null_on_error(date_formatter), x), coerce_to_list, self.resolve_property)
else:
fn = self.resolve_property
return (property, (
'//dc:{}/node()'.format(property),
'//ns0:{}/node()'.format(property),
fn)
)
def normalize(raw_doc, harvester_name):
normalized_started = timestamp()
harvester = registry[harvester_name]
normalized = null_on_error(harvester.normalize)(raw_doc)
if not normalized:
raise events.Skip('Did not normalize document with id {}'.format(raw_doc['docID']))
normalized['timestamps'] = util.stamp_from_raw(raw_doc, normalizeStarted=normalized_started)
return normalized # returns a single normalized document
def normalize(raw_doc, harvester_name):
normalized_started = timestamp()
harvester = registry[harvester_name]
normalized = null_on_error(harvester.normalize)(raw_doc)
if not normalized:
raise events.Skip('Did not normalize document with id {}'.format(
raw_doc['docID']))
normalized['timestamps'] = util.stamp_from_raw(
raw_doc, normalizeStarted=normalized_started)
return normalized # returns a single normalized document
