def from_ova(
cls, obj: ova.Analysis, nlp: t.Optional[t.Callable[[str], t.Any]]
) -> Resource:
return cls(
utils.parse(obj.get("plain_txt"), nlp),
obj.get("documentTitle"),
obj.get("documentSource"),
dt.from_format(obj.get("documentDate"), ova.DATE_FORMAT_ANALYSIS),
)
def from_sadface(
cls,
obj: sadface.Node,
nlp: t.Optional[t.Callable[[str], t.Any]] = None,
) -> AtomNode:
"""Generate AtomNode object from SADFace Node object."""
timestamp = pendulum.now()
return cls(
id=obj["id"],
text=utils.parse(obj["text"], nlp),
userdata=obj["metadata"],
metadata=Metadata(timestamp, timestamp),
)
def from_ova(
cls,
obj: ova.Node,
nlp: t.Optional[t.Callable[[str], t.Any]] = None,
) -> AtomNode:
"""Generate AtomNode object from OVA Node object."""
timestamp = dt.from_format(obj.get("date"),
ova.DATE_FORMAT) or pendulum.now()
return cls(
id=str(obj["id"]),
metadata=Metadata(timestamp, timestamp),
text=utils.parse(obj["text"], nlp),
)
def from_aif(
cls,
obj: aif.Node,
nlp: t.Optional[t.Callable[[str], t.Any]] = None,
) -> AtomNode:
"""Generate AtomNode object from AIF Node object."""
timestamp = (dt.from_format(obj.get("timestamp"), aif.DATE_FORMAT)
or pendulum.now())
return cls(
id=obj["nodeID"],
metadata=Metadata(timestamp, timestamp),
text=utils.parse(obj["text"], nlp),
)
def from_protobuf(
cls,
obj: graph_pb2.Reference,
resources: t.Mapping[str, Resource],
nlp: t.Optional[t.Callable[[str], t.Any]] = None,
) -> t.Optional[Reference]:
"""Generate Resource object from PROTOBUF format Graph's Resource object."""
if obj.text:
if obj.resource:
return cls(
resources[obj.resource],
obj.offset,
utils.parse(obj.text, nlp),
)
else:
return cls(
None,
None,
utils.parse(obj.text, nlp),
)
return None
def from_protobuf(
cls,
id: str,
obj: graph_pb2.Resource,
nlp: t.Optional[t.Callable[[str], t.Any]] = None,
) -> Resource:
"""Generate Resource object from PROTOBUF format Graph's Resource object."""
return cls(
utils.parse(obj.text, nlp),
obj.title,
obj.source,
dt.from_protobuf(obj.timestamp),
Metadata.from_protobuf(obj.metadata),
dict(obj.userdata.items()),
id,
)
def from_protobuf(
cls,
id: str,
obj: graph_pb2.Node,
resources: t.Mapping[str, Resource],
participants: t.Mapping[str, Participant],
reference_class: t.Type[Reference],
nlp: t.Optional[t.Callable[[str], t.Any]] = None,
) -> AtomNode:
"""Generate AtomNode object from PROTOBUF Node object."""
return cls(
utils.parse(obj.atom.text, nlp),
reference_class.from_protobuf(obj.atom.reference, resources, nlp),
participants.get(obj.atom.participant),
Metadata.from_protobuf(obj.metadata),
dict(obj.userdata.items()),
id=id,
)
def from_aml(
cls,
obj: et.Element,
nlp: t.Optional[t.Callable[[str], t.Any]] = None,
) -> AtomNode:
"""
Generate Node object from AML Node format. obj is a AML "PROP" element.
"""
# get id of PROP
if "identifier" in obj.attrib:
id = obj.get("identifier")
else:
id = None
# read text of PROP
text = obj.find("PROPTEXT").text
# read owners of PROP
owner_list = obj.findall("OWNER")
owners_lst = []
if not owner_list:
# if not empty, do something
for owner in owner_list:
owners_lst.append(owner.get("name"))
owners = {"owners": ", ".join(owners_lst)}
else:
owners = {}
# create timestamp
timestamp = pendulum.now()
return cls(
id=id,
text=utils.parse(text, nlp),
metadata=Metadata(timestamp, timestamp),
userdata=owners,
)
