def post(self, request, resourceid=None):
try:
indent = int(request.POST.get('indent', None))
except:
indent = None
try:
if user_can_edit_resources(user=request.user):
data = JSONDeserializer().deserialize(request.body)
reader = JsonLdReader()
reader.read_resource(data)
if reader.errors:
response = []
for value in reader.errors.itervalues():
response.append(value.message)
return JSONResponse(data, indent=indent, status=400, reason=response)
else:
response = []
for resource in reader.resources:
with transaction.atomic():
resource.save(request=request)
response.append(JSONDeserializer().deserialize(
self.get(request, resource.resourceinstanceid).content))
return JSONResponse(response, indent=indent)
else:
return JSONResponse(status=403)
except Exception as e:
return JSONResponse(status=500, reason=e)
def test_find_leaf_branch(self):
"""
Given a list of leaf nodes, find the appropriate node from the given jsonld
"""
jsonld_graph = {
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by": {
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E82_Actor_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value": "Will",
}
}
graphtree = self.unique_graph.get_tree()
for child in graphtree["children"]:
if child["node"].name == "Name":
node = child
reader = JsonLdReader()
branch = reader.findBranch(
node["children"],
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by",
jsonld_graph[
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by"],
)
self.assertEqual(str(branch["node"].pk),
"839b0e4c-95e6-11e8-aada-14109fd34195")
def post(self, request, resourceid=None, slug=None, graphid=None):
try:
indent = int(request.POST.get("indent", None))
except Exception:
indent = None
try:
if user_can_edit_resources(user=request.user):
data = JSONDeserializer().deserialize(request.body)
reader = JsonLdReader()
if slug is not None:
graphid = models.GraphModel.objects.get(slug=slug).pk
reader.read_resource(data, graphid=graphid)
if reader.errors:
response = []
for value in reader.errors.values():
response.append(value.message)
return JSONResponse({"error": response}, indent=indent, status=400)
else:
response = []
for resource in reader.resources:
with transaction.atomic():
resource.save(request=request)
response.append(JSONDeserializer().deserialize(self.get(request, resource.resourceinstanceid).content))
return JSONResponse(response, indent=indent, status=201)
else:
return JSONResponse(status=403)
except Exception as e:
if settings.DEBUG is True:
exc_type, exc_value, exc_traceback = sys.exc_info()
formatted = traceback.format_exception(exc_type, exc_value, exc_traceback)
if len(formatted):
for message in formatted:
print(message)
return JSONResponse({"error": "resource data could not be saved: %s" % e}, status=500, reason=e)
def put(self, request, resourceid):
if user_can_edit_resources(user=request.user):
data = JSONDeserializer().deserialize(request.body)
#print data
reader = JsonLdReader()
reader.read_resource(data)
else:
return JSONResponse(status=500)
return JSONResponse(self.get(request, resourceid))
def put(self, request, resourceid, slug=None, graphid=None):
try:
indent = int(request.PUT.get("indent", None))
except Exception:
indent = None
if not user_can_edit_resources(user=request.user,
resourceid=resourceid):
return JSONResponse(status=403)
else:
with transaction.atomic():
try:
# DELETE
resource_instance = Resource.objects.get(pk=resourceid)
resource_instance.delete()
except models.ResourceInstance.DoesNotExist:
pass
try:
# POST
data = JSONDeserializer().deserialize(request.body)
reader = JsonLdReader()
if slug is not None:
graphid = models.GraphModel.objects.get(slug=slug).pk
reader.read_resource(data,
resourceid=resourceid,
graphid=graphid)
if reader.errors:
response = []
for value in reader.errors.values():
response.append(value.message)
return JSONResponse({"error": response},
indent=indent,
status=400)
else:
response = []
for resource in reader.resources:
with transaction.atomic():
resource.save(request=request)
response.append(JSONDeserializer().deserialize(
self.get(request,
resource.resourceinstanceid).content))
return JSONResponse(response,
indent=indent,
status=201)
except models.ResourceInstance.DoesNotExist:
return JSONResponse(status=404)
except Exception as e:
return JSONResponse(
{"error": "resource data could not be saved"},
status=500,
reason=e)
def put(self, request, resourceid):
try:
indent = int(request.POST.get('indent', None))
except:
indent = None
try:
if user_can_edit_resources(user=request.user):
data = JSONDeserializer().deserialize(request.body)
reader = JsonLdReader()
reader.read_resource(data, use_ids=True)
if reader.errors:
response = []
for value in reader.errors.itervalues():
response.append(value.message)
return JSONResponse(data,
indent=indent,
status=400,
reason=response)
else:
response = []
for resource in reader.resources:
if resourceid != str(resource.pk):
raise Exception(
'Resource id in the URI does not match the resource @id supplied in the document'
)
old_resource = Resource.objects.get(pk=resource.pk)
old_resource.load_tiles()
old_tile_ids = set(
[str(tile.pk) for tile in old_resource.tiles])
new_tile_ids = set([
str(tile.pk)
for tile in resource.get_flattened_tiles()
])
tileids_to_delete = old_tile_ids.difference(
new_tile_ids)
tiles_to_delete = models.TileModel.objects.filter(
pk__in=tileids_to_delete)
with transaction.atomic():
tiles_to_delete.delete()
resource.save(request=request)
response.append(JSONDeserializer().deserialize(
self.get(request,
resource.resourceinstanceid).content))
return JSONResponse(response, indent=indent)
else:
return JSONResponse(status=403)
except Exception as e:
return JSONResponse(status=500, reason=e)
def test_find_branch_from_jsonld_2(self):
"""
The same test as above except that we now add an additional node to the supplied json which now will match a branch in the graph
The graph is partially unique (the children of the root are not unique)
"""
ambiguous_jsonld_graph = {
"@id":
"http://*****:*****@type": [
"http://www.cidoc-crm.org/cidoc-crm/E82_Actor_Appellation",
"http://*****:*****@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E41_Appellation",
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by": {
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E82_Actor_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value": "Will",
},
"http://www.cidoc-crm.org/cidoc-crm/P1i_identifies": {
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E41_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"Smith",
},
}],
}
graphtree = self.ambiguous_graph.get_tree()
reader = JsonLdReader()
branch = reader.findBranch(
graphtree["children"],
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by",
ambiguous_jsonld_graph[
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by"],
)
self.assertEqual(str(branch["node"].pk),
"3f40c4c0-9693-11e8-8a0f-14109fd34195")
def test_find_unique_branch_from_jsonld(self):
"""
Test that we can find the correct branch in the graph that matches the supplied json-ld
The graph is partially unique (the children of the root are not unique)
"""
jsonld_graph = {
"@id":
"http://*****:*****@type": [
"http://*****:*****@id":
"http://*****:*****@type": "http://www.cidoc-crm.org/cidoc-crm/E41_Appellation",
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by": {
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E82_Actor_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value": "Will",
},
"http://www.cidoc-crm.org/cidoc-crm/P1i_identifies": {
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E41_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"Smith",
},
},
}
graphtree = self.unique_graph.get_tree()
reader = JsonLdReader()
branch = reader.findBranch(
graphtree["children"],
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by",
jsonld_graph[
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by"],
)
self.assertEqual(str(branch["node"].pk),
"3e1e65dc-95e6-11e8-9de9-14109fd34195")
def put(self, request, resourceid):
try:
indent = int(request.PUT.get('indent', None))
except:
indent = None
if user_can_edit_resources(user=request.user):
with transaction.atomic():
try:
# DELETE
resource_instance = Resource.objects.get(pk=resourceid)
resource_instance.delete()
except models.ResourceInstance.DoesNotExist:
pass
try:
# POST
data = JSONDeserializer().deserialize(request.body)
reader = JsonLdReader()
reader.read_resource(data, resourceid=resourceid)
if reader.errors:
response = []
for value in reader.errors.itervalues():
response.append(value.message)
return JSONResponse(data,
indent=indent,
status=400,
reason=response)
else:
response = []
for resource in reader.resources:
with transaction.atomic():
resource.save(request=request)
response.append(JSONDeserializer().deserialize(
self.get(request,
resource.resourceinstanceid).content))
return JSONResponse(response,
indent=indent,
status=201)
except models.ResourceInstance.DoesNotExist:
return JSONResponse(status=404)
else:
return JSONResponse(status=500)
def test_cant_find_branch_from_ambiguous_jsonld(self):
"""
Test that we raise the appropriate error when we can't find the correct branch in the graph given that the supplied
json-ld could match more than one branch
The graph is partially unique (the children of the root are not unique)
"""
ambiguous_jsonld_graph = {
"@id":
"http://*****:*****@type": [
"http://www.cidoc-crm.org/cidoc-crm/E82_Actor_Appellation",
"http://*****:*****@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E41_Appellation",
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by": {
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E82_Actor_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"Will - Ambiguous",
},
}],
}
graphtree = self.ambiguous_graph.get_tree()
reader = JsonLdReader()
with self.assertRaises(reader.AmbiguousGraphException) as cm:
branch = reader.findBranch(
graphtree["children"],
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by",
ambiguous_jsonld_graph[
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by"],
)
def test_find_other_unique_branch_from_jsonld(self):
"""
Test that we can find the correct branch in the graph that matches the supplied json-ld
The graph is partially unique (the children of the root are not unique)
"""
jsonld_graph = {
"@id":
"http://*****:*****@type": [
"http://www.cidoc-crm.org/cidoc-crm/E82_Actor_Appellation",
"http://*****:*****@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E41_Appellation",
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by": {
"@id":
"http://*****:*****@type":
"http://www.ics.forth.gr/isl/CRMdig/D21_Person_Name",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"The Shadow",
},
}],
}
graphtree = self.unique_graph.get_tree()
reader = JsonLdReader()
branch = reader.findBranch(
graphtree["children"],
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by",
jsonld_graph[
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by"],
)
self.assertEqual(str(branch["node"].pk),
"91679e1e-95e6-11e8-a166-14109fd34195")
def test_cant_find_branch_from_jsonld(self):
"""
Test that we raise the appropriate error when we can't find the correct branch in the graph that matches the supplied json-ld
The graph is partially unique (the children of the root are not unique)
"""
incorrect_jsonld_graph = {
"@id":
"http://*****:*****@type": [
"http://www.cidoc-crm.org/cidoc-crm/E82_Actor_Appellation",
"http://*****:*****@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E41_Appellation",
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by": {
"@id":
"http://*****:*****@type":
"---THIS TYPE IS INCORRECT AND SHOULN'T MATCH---",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"The Shadow",
},
}],
}
graphtree = self.unique_graph.get_tree()
reader = JsonLdReader()
with self.assertRaises(reader.DataDoesNotMatchGraphException) as cm:
branch = reader.findBranch(
graphtree["children"],
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by",
incorrect_jsonld_graph[
"http://www.cidoc-crm.org/cidoc-crm/P1_is_identified_by"],
)
def test_cant_find_branch_from_complex_ambigious_jsonld(self):
"""
The same test as above except that we now supply a jsonld structure that matches more then one branch in the graph (it's ambiguous)
"""
complex_jsonld_graph = {
"@id":
"http://*****:*****@type": [
"http://www.cidoc-crm.org/cidoc-crm/E12_Production",
"http://*****:*****@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E17_Type_Assignment",
"http://www.cidoc-crm.org/cidoc-crm/P42_assigned": [
{
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E55_Type",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"174e9486-0663-4c9d-ab78-c7e441720c26",
},
{
"@id":
"http://*****:*****@type": "http://www.cidoc-crm.org/cidoc-crm/E55_Type",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"None",
},
],
"http://www.cidoc-crm.org/cidoc-crm/P4_has_time-span": {
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E52_Time-Span",
"http://www.cidoc-crm.org/cidoc-crm/P78_is_identified_by":
[
{
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E49_Time_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"2018-08-06",
},
{
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E49_Time_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"2018-09-20",
},
],
},
}],
}
graphtree = self.phase_type_assignment_graph.get_tree()
reader = JsonLdReader()
with self.assertRaises(reader.AmbiguousGraphException) as cm:
branch = reader.findBranch(
graphtree["children"],
"http://www.cidoc-crm.org/cidoc-crm/P41i_was_classified_by",
complex_jsonld_graph[
"http://www.cidoc-crm.org/cidoc-crm/P41i_was_classified_by"],
)
def test_find_branch_from_complex_jsonld(self):
"""
Given a more complicated json structure find the branch in the graph
"""
complex_jsonld_graph = {
"@id":
"http://*****:*****@type": [
"http://www.cidoc-crm.org/cidoc-crm/E12_Production",
"http://*****:*****@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E17_Type_Assignment",
"http://www.cidoc-crm.org/cidoc-crm/P42_assigned": {
"@id":
"http://*****:*****@type": "http://www.cidoc-crm.org/cidoc-crm/E55_Type",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value": "None",
},
"http://www.cidoc-crm.org/cidoc-crm/P2_has_type": {
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E55_Type",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"51cbfba6-34ee-4fbd-8b6e-10ef73fd4083",
},
"http://www.cidoc-crm.org/cidoc-crm/P4_has_time-span": {
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E52_Time-Span",
"http://www.cidoc-crm.org/cidoc-crm/P78_is_identified_by":
[
{
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E49_Time_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"2018-08-05",
},
{
"@id":
"http://*****:*****@type":
"http://www.cidoc-crm.org/cidoc-crm/E49_Time_Appellation",
"http://www.w3.org/1999/02/22-rdf-syntax-ns#value":
"2018-08-06",
},
],
},
}],
}
graphtree = self.phase_type_assignment_graph.get_tree()
reader = JsonLdReader()
branch = reader.findBranch(
graphtree["children"],
"http://www.cidoc-crm.org/cidoc-crm/P41i_was_classified_by",
complex_jsonld_graph[
"http://www.cidoc-crm.org/cidoc-crm/P41i_was_classified_by"],
)
self.assertEqual(str(branch["node"].pk),
"049fc0c9-fa36-11e6-9e3e-026d961c88e6")
class Command(BaseCommand):
"""
Command for importing JSON-LD data into Arches
"""
def add_arguments(self, parser):
parser.add_argument(
"-s", "--source", default="data/", action="store", dest="source", help="the directory in which the data files are to be found"
)
parser.add_argument(
"-ow",
"--overwrite",
default="ignore",
action="store",
dest="force",
help="if overwrite, overwrite records that exist; if ignore, then skip; if error, then halt",
)
parser.add_argument("--toobig", default=0, type=int, action="store", dest="toobig", help="Do not attempt to load records > n kb")
parser.add_argument(
"-m", "--model", default="", action="store", dest="model", help="the name of the model path to load (eg auction_of_lot)",
)
parser.add_argument(
"-b",
"--block",
default="",
action="store",
dest="block",
help="the name of the block in the model path to load (eg 00), or slice in the form this,total (eg 1,5)",
)
parser.add_argument("--max", default=-1, type=int, action="store", dest="max", help="Maximum number of records to load per model")
parser.add_argument("--fast", default=0, action="store", type=int, dest="fast", help="Use bulk_save to store n records at a time")
parser.add_argument("-q", "--quiet", default=False, action="store_true", dest="quiet", help="Don't announce every record")
parser.add_argument(
"--skip", default=-1, type=int, action="store", dest="skip", help="Number of records to skip before starting to load"
)
parser.add_argument("--suffix", default="json", action="store", dest="suffix", help="file suffix to load if not .json")
parser.add_argument(
"--ignore-errors", default=False, action="store_true", dest="ignore_errors", help="Log but do not terminate on errors"
)
parser.add_argument(
"--strip-issearchable",
default=False,
action="store_true",
dest="strip_search",
help="If a node is set to not be exposed to advanced search, then don't even index it",
)
def handle(self, *args, **options):
print("Starting JSON-LD load")
if options["model"]:
print(f"Only loading {options['model']}")
if options["block"]:
print(f"Only loading {options['block']}")
if options["force"] == "overwrite":
print("Overwriting existing records")
if options["toobig"]:
print(f"Not loading records > {options['toobig']}kb")
if options["quiet"]:
print("Only announcing timing data")
if options["strip_search"] and not options["fast"]:
print("ERROR: stripping fields not exposed to advanced search only works in fast mode")
return
self.resources = []
self.load_resources(options)
def load_resources(self, options):
self.reader = JsonLdReader()
self.jss = JSONSerializer()
source = options["source"]
if options["model"]:
models = [options["model"]]
else:
models = os.listdir(source)
models.sort()
models = [m for m in models if m[0] not in ["_", "."]]
print(f"Found possible models: {models}")
# This is boilerplate for any use of get_documents_to_index()
# Need to add issearchable for strip_search option
# Only calculate it once per load
self.datatype_factory = DataTypeFactory()
dt_instance_hash = {}
self.node_info = {
str(nodeid): {
"datatype": dt_instance_hash.setdefault(datatype, self.datatype_factory.get_instance(datatype)),
"issearchable": srch,
}
for nodeid, datatype, srch in archesmodels.Node.objects.values_list("nodeid", "datatype", "issearchable")
}
self.node_datatypes = {str(nodeid): datatype for nodeid, datatype in archesmodels.Node.objects.values_list("nodeid", "datatype")}
start = time.time()
seen = 0
loaded = 0
for m in models:
print(f"Loading {m}")
graphid = graph_uuid_map.get(m, None)
if not graphid:
# Check slug
try:
graphid = archesmodels.GraphModel.objects.get(slug=m).pk
except:
print(f"Couldn't find a model definition for {m}; skipping")
continue
# We have a good model, so build the pre-processed tree once
self.reader.graphtree = self.reader.process_graph(graphid)
block = options["block"]
if block and "," not in block:
blocks = [block]
else:
blocks = os.listdir(f"{source}/{m}")
blocks.sort()
blocks = [b for b in blocks if b[0] not in ["_", "."]]
if "," in block:
# {slice},{max-slices}
(cslice, mslice) = block.split(",")
cslice = int(cslice) - 1
mslice = int(mslice)
blocks = blocks[cslice::mslice]
loaded_model = 0
try:
for b in blocks:
files = os.listdir(f"{source}/{m}/{b}")
files.sort()
for f in files:
if not f.endswith(options["suffix"]):
continue
elif f.startswith(".") or f.startswith("_"):
continue
if options["max"] > 0 and loaded_model >= options["max"]:
raise StopIteration()
seen += 1
if seen <= options["skip"]:
# Do it this way to keep the counts correct
continue
fn = f"{source}/{m}/{b}/{f}"
# Check file size of record
if not options["quiet"]:
print(f"About to import {fn}")
if options["toobig"]:
sz = os.os.path.getsize(fn)
if sz > options["toobig"]:
if not quiet:
print(f" ... Skipping due to size: {sz} > {options['toobig']}")
continue
uu = f.replace(f".{options['suffix']}", "")
fh = open(fn)
data = fh.read()
fh.close()
# FIXME Timezone / DateTime Workaround
# FIXME The following line should be removed when #5669 / #6346 are closed
data = data.replace("T00:00:00Z", "")
jsdata = json.loads(data)
jsdata = fix_js_data(data, jsdata, m)
if len(uu) != 36 or uu[8] != "-":
# extract uuid from data if filename is not a UUID
uu = jsdata["id"][-36:]
if jsdata:
try:
if options["fast"]:
l = self.fast_import_resource(
uu,
graphid,
jsdata,
n=options["fast"],
reload=options["force"],
quiet=options["quiet"],
strip_search=options["strip_search"],
)
else:
l = self.import_resource(uu, graphid, jsdata, reload=options["force"], quiet=options["quiet"])
loaded += l
loaded_model += l
except Exception as e:
print(f"*** Failed to load {fn}:\n {e}\n")
if not options["ignore_errors"]:
raise
else:
print(" ... skipped due to bad data :(")
if not seen % 100:
print(f" ... seen {seen} / loaded {loaded} in {time.time()-start}")
except StopIteration as e:
break
except:
raise
if options["fast"] and self.resources:
self.save_resources()
self.index_resources(options["strip_search"])
self.resources = []
print(f"Total Time: seen {seen} / loaded {loaded} in {time.time()-start} seconds")
def fast_import_resource(self, resourceid, graphid, data, n=1000, reload="ignore", quiet=True, strip_search=False):
try:
resource_instance = Resource.objects.get(pk=resourceid)
if reload == "ignore":
if not quiet:
print(f" ... already loaded")
return 0
elif reload == "error":
print(f"*** Record exists for {resourceid}, and -ow is error")
raise FileExistsError(resourceid)
else:
resource_instance.delete()
except archesmodels.ResourceInstance.DoesNotExist:
# thrown when resource doesn't exist
pass
try:
self.reader.read_resource(data, resourceid=resourceid, graphid=graphid)
self.resources.extend(self.reader.resources)
except:
print(f"Exception raised while reading {resourceid}...")
raise
if len(self.resources) >= n:
self.save_resources()
self.index_resources(strip_search)
self.resources = []
return 1
def import_resource(self, resourceid, graphid, data, reload="ignore", quiet=False):
with transaction.atomic():
try:
resource_instance = Resource.objects.get(pk=resourceid)
if reload == "ignore":
if not quiet:
print(f" ... already loaded")
return 0
elif reload == "error":
print(f"*** Record exists for {resourceid}, and -ow is error")
raise FileExistsError(resourceid)
else:
resource_instance.delete()
except archesmodels.ResourceInstance.DoesNotExist:
# thrown when resource doesn't exist
pass
try:
self.reader.read_resource(data, resourceid=resourceid, graphid=graphid)
for resource in self.reader.resources:
resource.save(request=None)
except archesmodels.ResourceInstance.DoesNotExist:
print(f"*** Could not find model: {graphid}")
return 0
except Exception as e:
raise
return 1
def save_resources(self):
tiles = []
for resource in self.resources:
resource.tiles = resource.get_flattened_tiles()
tiles.extend(resource.tiles)
Resource.objects.bulk_create(self.resources)
TileModel.objects.bulk_create(tiles)
for t in tiles:
for nodeid in t.data.keys():
datatype = self.node_info[nodeid]["datatype"]
datatype.pre_tile_save(t, nodeid)
for resource in self.resources:
resource.save_edit(edit_type="create")
def index_resources(self, strip_search=False):
se = SearchEngineInstance
documents = []
term_list = []
for resource in self.resources:
if strip_search:
document, terms = monkey_get_documents_to_index(resource, node_info=self.node_info)
else:
document, terms = resource.get_documents_to_index(
fetchTiles=False, datatype_factory=self.datatype_factory, node_datatypes=self.node_datatypes
)
documents.append(se.create_bulk_item(index="resources", id=document["resourceinstanceid"], data=document))
for term in terms:
term_list.append(se.create_bulk_item(index="terms", id=term["_id"], data=term["_source"]))
se.bulk_index(documents)
se.bulk_index(term_list)
def load_resources(self, options):
self.reader = JsonLdReader()
self.jss = JSONSerializer()
source = options["source"]
if options["model"]:
models = [options["model"]]
else:
models = os.listdir(source)
models.sort()
models = [m for m in models if m[0] not in ["_", "."]]
print(f"Found possible models: {models}")
# This is boilerplate for any use of get_documents_to_index()
# Need to add issearchable for strip_search option
# Only calculate it once per load
self.datatype_factory = DataTypeFactory()
dt_instance_hash = {}
self.node_info = {
str(nodeid): {
"datatype": dt_instance_hash.setdefault(datatype, self.datatype_factory.get_instance(datatype)),
"issearchable": srch,
}
for nodeid, datatype, srch in archesmodels.Node.objects.values_list("nodeid", "datatype", "issearchable")
}
self.node_datatypes = {str(nodeid): datatype for nodeid, datatype in archesmodels.Node.objects.values_list("nodeid", "datatype")}
start = time.time()
seen = 0
loaded = 0
for m in models:
print(f"Loading {m}")
graphid = graph_uuid_map.get(m, None)
if not graphid:
# Check slug
try:
graphid = archesmodels.GraphModel.objects.get(slug=m).pk
except:
print(f"Couldn't find a model definition for {m}; skipping")
continue
# We have a good model, so build the pre-processed tree once
self.reader.graphtree = self.reader.process_graph(graphid)
block = options["block"]
if block and "," not in block:
blocks = [block]
else:
blocks = os.listdir(f"{source}/{m}")
blocks.sort()
blocks = [b for b in blocks if b[0] not in ["_", "."]]
if "," in block:
# {slice},{max-slices}
(cslice, mslice) = block.split(",")
cslice = int(cslice) - 1
mslice = int(mslice)
blocks = blocks[cslice::mslice]
loaded_model = 0
try:
for b in blocks:
files = os.listdir(f"{source}/{m}/{b}")
files.sort()
for f in files:
if not f.endswith(options["suffix"]):
continue
elif f.startswith(".") or f.startswith("_"):
continue
if options["max"] > 0 and loaded_model >= options["max"]:
raise StopIteration()
seen += 1
if seen <= options["skip"]:
# Do it this way to keep the counts correct
continue
fn = f"{source}/{m}/{b}/{f}"
# Check file size of record
if not options["quiet"]:
print(f"About to import {fn}")
if options["toobig"]:
sz = os.os.path.getsize(fn)
if sz > options["toobig"]:
if not quiet:
print(f" ... Skipping due to size: {sz} > {options['toobig']}")
continue
uu = f.replace(f".{options['suffix']}", "")
fh = open(fn)
data = fh.read()
fh.close()
# FIXME Timezone / DateTime Workaround
# FIXME The following line should be removed when #5669 / #6346 are closed
data = data.replace("T00:00:00Z", "")
jsdata = json.loads(data)
jsdata = fix_js_data(data, jsdata, m)
if len(uu) != 36 or uu[8] != "-":
# extract uuid from data if filename is not a UUID
uu = jsdata["id"][-36:]
if jsdata:
try:
if options["fast"]:
l = self.fast_import_resource(
uu,
graphid,
jsdata,
n=options["fast"],
reload=options["force"],
quiet=options["quiet"],
strip_search=options["strip_search"],
)
else:
l = self.import_resource(uu, graphid, jsdata, reload=options["force"], quiet=options["quiet"])
loaded += l
loaded_model += l
except Exception as e:
print(f"*** Failed to load {fn}:\n {e}\n")
if not options["ignore_errors"]:
raise
else:
print(" ... skipped due to bad data :(")
if not seen % 100:
print(f" ... seen {seen} / loaded {loaded} in {time.time()-start}")
except StopIteration as e:
break
except:
raise
if options["fast"] and self.resources:
self.save_resources()
self.index_resources(options["strip_search"])
self.resources = []
print(f"Total Time: seen {seen} / loaded {loaded} in {time.time()-start} seconds")
