def append_dsl(self, search_results_object, permitted_nodegroups,
include_provisional):
nested_agg = NestedAgg(path='points', name='geo_aggs')
nested_agg_filter = FiltersAgg(name='inner')
geo_agg_filter = Bool()
if include_provisional is True:
geo_agg_filter.filter(
Terms(field='points.provisional', terms=['false', 'true']))
else:
if include_provisional is False:
geo_agg_filter.filter(
Terms(field='points.provisional', terms=['false']))
elif include_provisional is 'only provisional':
geo_agg_filter.filter(
Terms(field='points.provisional', terms=['true']))
geo_agg_filter.filter(
Terms(field='points.nodegroup_id', terms=permitted_nodegroups))
nested_agg_filter.add_filter(geo_agg_filter)
nested_agg_filter.add_aggregation(
GeoHashGridAgg(field='points.point',
name='grid',
precision=settings.HEX_BIN_PRECISION))
nested_agg_filter.add_aggregation(
GeoBoundsAgg(field='points.point', name='bounds'))
nested_agg.add_aggregation(nested_agg_filter)
search_results_object['query'].add_aggregation(nested_agg)
def append_dsl(self, search_results_object, permitted_nodegroups,
include_provisional):
nested_agg = NestedAgg(path="points", name="geo_aggs")
nested_agg_filter = FiltersAgg(name="inner")
geo_agg_filter = Bool()
if include_provisional is True:
geo_agg_filter.filter(
Terms(field="points.provisional", terms=["false", "true"]))
else:
if include_provisional is False:
geo_agg_filter.filter(
Terms(field="points.provisional", terms=["false"]))
elif include_provisional is "only provisional":
geo_agg_filter.filter(
Terms(field="points.provisional", terms=["true"]))
geo_agg_filter.filter(
Terms(field="points.nodegroup_id", terms=permitted_nodegroups))
nested_agg_filter.add_filter(geo_agg_filter)
nested_agg_filter.add_aggregation(
GeoHashGridAgg(field="points.point",
name="grid",
precision=settings.HEX_BIN_PRECISION))
nested_agg_filter.add_aggregation(
GeoBoundsAgg(field="points.point", name="bounds"))
nested_agg.add_aggregation(nested_agg_filter)
search_results_object["query"].add_aggregation(nested_agg)
def get_resource_bounds(node):
query = Query(se, start=0, limit=0)
search_query = Bool()
query.add_query(search_query)
query.add_aggregation(
GeoBoundsAgg(field='points.point', name='bounds'))
results = query.search(index='resource',
doc_type=[str(node.graph_id)])
bounds = results['aggregations']['bounds'][
'bounds'] if 'bounds' in results['aggregations'][
'bounds'] else None
return bounds
def get_resource_bounds(node):
query = Query(se, start=0, limit=0)
search_query = Bool()
query.add_query(search_query)
query.add_aggregation(
GeoBoundsAgg(field="points.point", name="bounds"))
query.add_query(
Term(field="graph_id", term=str(node.graph.graphid)))
results = query.search(index=RESOURCES_INDEX)
bounds = results["aggregations"]["bounds"][
"bounds"] if "bounds" in results["aggregations"][
"bounds"] else None
return bounds
def build_search_results_dsl(request):
term_filter = request.GET.get('termFilter', '')
spatial_filter = JSONDeserializer().deserialize(
request.GET.get('mapFilter', '{}'))
include_provisional = get_provisional_type(request)
export = request.GET.get('export', None)
mobile_download = request.GET.get('mobiledownload', None)
page = 1 if request.GET.get('page') == '' else int(
request.GET.get('page', 1))
temporal_filter = JSONDeserializer().deserialize(
request.GET.get('temporalFilter', '{}'))
advanced_filters = JSONDeserializer().deserialize(
request.GET.get('advanced', '[]'))
search_buffer = None
se = SearchEngineFactory().create()
if export != None:
limit = settings.SEARCH_EXPORT_ITEMS_PER_PAGE
elif mobile_download != None:
limit = settings.MOBILE_DOWNLOAD_RESOURCE_LIMIT
else:
limit = settings.SEARCH_ITEMS_PER_PAGE
query = Query(se, start=limit * int(page - 1), limit=limit)
search_query = Bool()
nested_agg = NestedAgg(path='points', name='geo_aggs')
nested_agg_filter = FiltersAgg(name='inner')
if include_provisional == True:
nested_agg_filter.add_filter(
Terms(field='points.provisional', terms=['false', 'true']))
else:
provisional_resource_filter = Bool()
if include_provisional == False:
provisional_resource_filter.filter(
Terms(field='provisional', terms=['false', 'partial']))
nested_agg_filter.add_filter(
Terms(field='points.provisional', terms=['false']))
elif include_provisional == 'only provisional':
provisional_resource_filter.filter(
Terms(field='provisional', terms=['true', 'partial']))
nested_agg_filter.add_filter(
Terms(field='points.provisional', terms=['true']))
search_query.must(provisional_resource_filter)
nested_agg_filter.add_aggregation(
GeoHashGridAgg(field='points.point',
name='grid',
precision=settings.HEX_BIN_PRECISION))
nested_agg_filter.add_aggregation(
GeoBoundsAgg(field='points.point', name='bounds'))
nested_agg.add_aggregation(nested_agg_filter)
query.add_aggregation(nested_agg)
permitted_nodegroups = get_permitted_nodegroups(request.user)
if term_filter != '':
for term in JSONDeserializer().deserialize(term_filter):
term_query = Bool()
provisional_term_filter = Bool()
if term['type'] == 'term' or term['type'] == 'string':
string_filter = Bool()
if term['type'] == 'term':
string_filter.must(
Match(field='strings.string',
query=term['value'],
type='phrase'))
elif term['type'] == 'string':
string_filter.should(
Match(field='strings.string',
query=term['value'],
type='phrase_prefix'))
string_filter.should(
Match(field='strings.string.folded',
query=term['value'],
type='phrase_prefix'))
if include_provisional == False:
string_filter.must_not(
Match(field='strings.provisional',
query='true',
type='phrase'))
elif include_provisional == 'only provisional':
string_filter.must_not(
Match(field='strings.provisional',
query='false',
type='phrase'))
string_filter.filter(
Terms(field='strings.nodegroup_id',
terms=permitted_nodegroups))
nested_string_filter = Nested(path='strings',
query=string_filter)
if term['inverted']:
search_query.must_not(nested_string_filter)
else:
search_query.must(nested_string_filter)
# need to set min_score because the query returns results with score 0 and those have to be removed, which I don't think it should be doing
query.min_score('0.01')
elif term['type'] == 'concept':
concept_ids = _get_child_concepts(term['value'])
conceptid_filter = Bool()
conceptid_filter.filter(
Terms(field='domains.conceptid', terms=concept_ids))
conceptid_filter.filter(
Terms(field='domains.nodegroup_id',
terms=permitted_nodegroups))
if include_provisional == False:
conceptid_filter.must_not(
Match(field='domains.provisional',
query='true',
type='phrase'))
elif include_provisional == 'only provisional':
conceptid_filter.must_not(
Match(field='domains.provisional',
query='false',
type='phrase'))
nested_conceptid_filter = Nested(path='domains',
query=conceptid_filter)
if term['inverted']:
search_query.must_not(nested_conceptid_filter)
else:
search_query.filter(nested_conceptid_filter)
if 'features' in spatial_filter:
if len(spatial_filter['features']) > 0:
feature_geom = spatial_filter['features'][0]['geometry']
feature_properties = {}
if 'properties' in spatial_filter['features'][0]:
feature_properties = spatial_filter['features'][0][
'properties']
buffer = {'width': 0, 'unit': 'ft'}
if 'buffer' in feature_properties:
buffer = feature_properties['buffer']
search_buffer = _buffer(feature_geom, buffer['width'],
buffer['unit'])
feature_geom = JSONDeserializer().deserialize(search_buffer.json)
geoshape = GeoShape(field='geometries.geom.features.geometry',
type=feature_geom['type'],
coordinates=feature_geom['coordinates'])
invert_spatial_search = False
if 'inverted' in feature_properties:
invert_spatial_search = feature_properties['inverted']
spatial_query = Bool()
if invert_spatial_search == True:
spatial_query.must_not(geoshape)
else:
spatial_query.filter(geoshape)
# get the nodegroup_ids that the user has permission to search
spatial_query.filter(
Terms(field='geometries.nodegroup_id',
terms=permitted_nodegroups))
if include_provisional == False:
spatial_query.filter(
Terms(field='geometries.provisional', terms=['false']))
elif include_provisional == 'only provisional':
spatial_query.filter(
Terms(field='geometries.provisional', terms=['true']))
search_query.filter(Nested(path='geometries', query=spatial_query))
if 'fromDate' in temporal_filter and 'toDate' in temporal_filter:
now = str(datetime.utcnow())
start_date = ExtendedDateFormat(temporal_filter['fromDate'])
end_date = ExtendedDateFormat(temporal_filter['toDate'])
date_nodeid = str(
temporal_filter['dateNodeId']
) if 'dateNodeId' in temporal_filter and temporal_filter[
'dateNodeId'] != '' else None
query_inverted = False if 'inverted' not in temporal_filter else temporal_filter[
'inverted']
temporal_query = Bool()
if query_inverted:
# inverted date searches need to use an OR clause and are generally more complicated to structure (can't use ES must_not)
# eg: less than START_DATE OR greater than END_DATE
inverted_date_query = Bool()
inverted_date_ranges_query = Bool()
if start_date.is_valid():
inverted_date_query.should(
Range(field='dates.date', lt=start_date.lower))
inverted_date_ranges_query.should(
Range(field='date_ranges.date_range', lt=start_date.lower))
if end_date.is_valid():
inverted_date_query.should(
Range(field='dates.date', gt=end_date.upper))
inverted_date_ranges_query.should(
Range(field='date_ranges.date_range', gt=end_date.upper))
date_query = Bool()
date_query.filter(inverted_date_query)
date_query.filter(
Terms(field='dates.nodegroup_id', terms=permitted_nodegroups))
if include_provisional == False:
date_query.filter(
Terms(field='dates.provisional', terms=['false']))
elif include_provisional == 'only provisional':
date_query.filter(
Terms(field='dates.provisional', terms=['true']))
if date_nodeid:
date_query.filter(Term(field='dates.nodeid', term=date_nodeid))
else:
date_ranges_query = Bool()
date_ranges_query.filter(inverted_date_ranges_query)
date_ranges_query.filter(
Terms(field='date_ranges.nodegroup_id',
terms=permitted_nodegroups))
if include_provisional == False:
date_ranges_query.filter(
Terms(field='date_ranges.provisional',
terms=['false']))
elif include_provisional == 'only provisional':
date_ranges_query.filter(
Terms(field='date_ranges.provisional', terms=['true']))
temporal_query.should(
Nested(path='date_ranges', query=date_ranges_query))
temporal_query.should(Nested(path='dates', query=date_query))
else:
date_query = Bool()
date_query.filter(
Range(field='dates.date',
gte=start_date.lower,
lte=end_date.upper))
date_query.filter(
Terms(field='dates.nodegroup_id', terms=permitted_nodegroups))
if include_provisional == False:
date_query.filter(
Terms(field='dates.provisional', terms=['false']))
elif include_provisional == 'only provisional':
date_query.filter(
Terms(field='dates.provisional', terms=['true']))
if date_nodeid:
date_query.filter(Term(field='dates.nodeid', term=date_nodeid))
else:
date_ranges_query = Bool()
date_ranges_query.filter(
Range(field='date_ranges.date_range',
gte=start_date.lower,
lte=end_date.upper,
relation='intersects'))
date_ranges_query.filter(
Terms(field='date_ranges.nodegroup_id',
terms=permitted_nodegroups))
if include_provisional == False:
date_ranges_query.filter(
Terms(field='date_ranges.provisional',
terms=['false']))
if include_provisional == 'only provisional':
date_ranges_query.filter(
Terms(field='date_ranges.provisional', terms=['true']))
temporal_query.should(
Nested(path='date_ranges', query=date_ranges_query))
temporal_query.should(Nested(path='dates', query=date_query))
search_query.filter(temporal_query)
datatype_factory = DataTypeFactory()
if len(advanced_filters) > 0:
advanced_query = Bool()
grouped_query = Bool()
grouped_queries = [grouped_query]
for index, advanced_filter in enumerate(advanced_filters):
tile_query = Bool()
for key, val in advanced_filter.iteritems():
if key != 'op':
node = models.Node.objects.get(pk=key)
if request.user.has_perm('read_nodegroup', node.nodegroup):
datatype = datatype_factory.get_instance(node.datatype)
datatype.append_search_filters(val, node, tile_query,
request)
nested_query = Nested(path='tiles', query=tile_query)
if advanced_filter['op'] == 'or' and index != 0:
grouped_query = Bool()
grouped_queries.append(grouped_query)
grouped_query.must(nested_query)
for grouped_query in grouped_queries:
advanced_query.should(grouped_query)
search_query.must(advanced_query)
query.add_query(search_query)
if search_buffer != None:
search_buffer = search_buffer.geojson
return {'query': query, 'search_buffer': search_buffer}
def build_search_results_dsl(request):
term_filter = request.GET.get('termFilter', '')
spatial_filter = JSONDeserializer().deserialize(
request.GET.get('mapFilter', '{}'))
export = request.GET.get('export', None)
page = 1 if request.GET.get('page') == '' else int(
request.GET.get('page', 1))
temporal_filter = JSONDeserializer().deserialize(
request.GET.get('temporalFilter', '{}'))
se = SearchEngineFactory().create()
if export != None:
limit = settings.SEARCH_EXPORT_ITEMS_PER_PAGE
else:
limit = settings.SEARCH_ITEMS_PER_PAGE
query = Query(se, start=limit * int(page - 1), limit=limit)
query.add_aggregation(
GeoHashGridAgg(field='points',
name='grid',
precision=settings.HEX_BIN_PRECISION))
query.add_aggregation(GeoBoundsAgg(field='points', name='bounds'))
search_query = Bool()
if term_filter != '':
for term in JSONDeserializer().deserialize(term_filter):
if term['type'] == 'term':
term_filter = Match(field='strings',
query=term['value'],
type='phrase')
if term['inverted']:
search_query.must_not(term_filter)
else:
search_query.must(term_filter)
elif term['type'] == 'concept':
concept_ids = _get_child_concepts(term['value'])
conceptid_filter = Terms(field='domains.conceptid',
terms=concept_ids)
if term['inverted']:
search_query.must_not(conceptid_filter)
else:
search_query.must(conceptid_filter)
elif term['type'] == 'string':
string_filter = Bool()
string_filter.should(
Match(field='strings',
query=term['value'],
type='phrase_prefix'))
string_filter.should(
Match(field='strings.folded',
query=term['value'],
type='phrase_prefix'))
if term['inverted']:
search_query.must_not(string_filter)
else:
search_query.must(string_filter)
if 'features' in spatial_filter:
if len(spatial_filter['features']) > 0:
feature_geom = spatial_filter['features'][0]['geometry']
feature_properties = spatial_filter['features'][0]['properties']
buffer = {'width': 0, 'unit': 'ft'}
if 'buffer' in feature_properties:
buffer = feature_properties['buffer']
feature_geom = JSONDeserializer().deserialize(
_buffer(feature_geom, buffer['width'], buffer['unit']).json)
geoshape = GeoShape(field='geometries.features.geometry',
type=feature_geom['type'],
coordinates=feature_geom['coordinates'])
invert_spatial_search = False
if 'inverted' in feature_properties:
invert_spatial_search = feature_properties['inverted']
if invert_spatial_search == True:
search_query.must_not(geoshape)
else:
search_query.must(geoshape)
if 'fromDate' in temporal_filter and 'toDate' in temporal_filter:
now = str(datetime.utcnow())
start_date = None
end_date = None
start_year = 'null'
end_year = 'null'
try:
# start_date = parser.parse(temporal_filter['fromDate'])
# start_date = start_date.isoformat()
sd = FlexiDate.from_str(temporal_filter['fromDate'])
start_date = int((sd.as_float() - 1970) * 31556952 * 1000)
#start_year = parser.parse(start_date).year
start_year = sd.year
except:
pass
try:
# end_date = parser.parse(temporal_filter['toDate'])
# end_date = end_date.isoformat()
ed = FlexiDate.from_str(temporal_filter['toDate'])
end_date = int((ed.as_float() - 1970) * 31556952 * 1000)
#end_year = parser.parse(end_date).year
end_year = ed.year
except:
pass
# add filter for concepts that define min or max dates
sql = None
basesql = """
SELECT value.conceptid
FROM (
SELECT
{select_clause},
v.conceptid
FROM
public."values" v,
public."values" v2
WHERE
v.conceptid = v2.conceptid and
v.valuetype = 'min_year' and
v2.valuetype = 'max_year'
) as value
WHERE overlap = true;
"""
temporal_query = Bool()
if 'inverted' not in temporal_filter:
temporal_filter['inverted'] = False
if temporal_filter['inverted']:
# inverted date searches need to use an OR clause and are generally more complicated to structure (can't use ES must_not)
# eg: less than START_DATE OR greater than END_DATE
select_clause = []
inverted_date_filter = Bool()
field = 'dates'
if 'dateNodeId' in temporal_filter and temporal_filter[
'dateNodeId'] != '':
field = 'tiles.data.%s' % (temporal_filter['dateNodeId'])
if start_date is not None:
inverted_date_filter.should(Range(field=field, lte=start_date))
select_clause.append(
"(numrange(v.value::int, v2.value::int, '[]') && numrange(null,{start_year},'[]'))"
)
if end_date is not None:
inverted_date_filter.should(Range(field=field, gte=end_date))
select_clause.append(
"(numrange(v.value::int, v2.value::int, '[]') && numrange({end_year},null,'[]'))"
)
if 'dateNodeId' in temporal_filter and temporal_filter[
'dateNodeId'] != '':
date_range_query = Nested(path='tiles',
query=inverted_date_filter)
temporal_query.should(date_range_query)
else:
temporal_query.should(inverted_date_filter)
select_clause = " or ".join(select_clause) + " as overlap"
sql = basesql.format(select_clause=select_clause).format(
start_year=start_year, end_year=end_year)
else:
if 'dateNodeId' in temporal_filter and temporal_filter[
'dateNodeId'] != '':
range = Range(field='tiles.data.%s' %
(temporal_filter['dateNodeId']),
gte=start_date,
lte=end_date)
date_range_query = Nested(path='tiles', query=range)
temporal_query.should(date_range_query)
else:
date_range_query = Range(field='dates',
gte=start_date,
lte=end_date)
temporal_query.should(date_range_query)
select_clause = """
numrange(v.value::int, v2.value::int, '[]') && numrange({start_year},{end_year},'[]') as overlap
"""
sql = basesql.format(select_clause=select_clause).format(
start_year=start_year, end_year=end_year)
# is a dateNodeId is not specified
if sql is not None:
cursor = connection.cursor()
cursor.execute(sql)
ret = [str(row[0]) for row in cursor.fetchall()]
if len(ret) > 0:
conceptid_filter = Terms(field='domains.conceptid', terms=ret)
temporal_query.should(conceptid_filter)
search_query.must(temporal_query)
query.add_query(search_query)
return query