class ExportJobSchema(Schema):
export_id = fields.Integer(attribute="export_job_id")
file_path = fields.String(attribute="export_file_path")
mark_type = fields.Function(lambda obj: NlpTaskEnum(obj.nlp_task_id).name) # nlp_task_id
export_state = fields.Function(lambda obj: StatusEnum(obj.export_job_status).name)
project_name = fields.String(attribute="doc_type_name")
created_time = fields.String()
mark_job_ids = fields.List(fields.Integer(), attribute="export_mark_job_ids")
class MarkTaskSchema(Schema): # type: ignore
task_id = fields.Integer(attribute="mark_task_id")
doc = fields.Nested(DocSchema)
doc_type = fields.Nested({
"doc_type_id": fields.Integer(),
"doc_type_name": fields.String(),
"doc_type_desc": fields.String(),
})
user_task_list = fields.List(fields.Nested(UserTaskSchema))
task_state = fields.Function(
lambda obj: StatusEnum(obj.mark_task_status).name)
status = fields.Function(lambda obj: not obj.is_deleted)
created_time = fields.String()
task_result = fields.List(fields.Dict, attribute="mark_task_result")
class UserTaskSchema(Schema):
doc = fields.Nested(DocSchema)
doc_type = fields.Nested({
"doc_type_id": fields.Integer(),
"doc_type_name": fields.String(),
"doc_type_desc": fields.String(),
})
task_id = fields.Integer(attribute="user_task_id")
labeler_id = fields.Integer(attribute="annotator_id")
manual_task_id = fields.Integer(attribute="mark_task_id")
task_result = fields.List(fields.Dict, attribute="user_task_result")
task_state = fields.Function(
lambda obj: StatusEnum(obj.user_task_status).name)
status = fields.Function(lambda obj: not obj.is_deleted)
created_time = fields.String()
class TrainTermTaskSchema(Schema): # type: ignore
train_term_id = fields.Integer(attribute="train_term_task_id")
train_term_result = fields.Dict()
train_term_state = fields.Function(
lambda obj: StatusEnum(obj.train_term_status).name)
model_train_id = fields.Integer(attribute="train_task_id")
doc_term_id = fields.Integer()
class PredictTaskSchema(Schema):
doc_type = fields.Nested(DocTypeSchema)
doc = fields.Nested(DocSchema)
task_id = fields.Integer(attribute="predict_task_id")
task_result = fields.Dict(attribute="predict_task_result")
task_state = fields.Function(
lambda obj: StatusEnum(obj.predict_task_status).name)
created_time = fields.DateTime()
class CustomAlgorithmSchema(Schema):
custom_id = fields.Integer(attribute="custom_algorithm_id")
custom_ip = fields.String(attribute="custom_algorithm_ip")
custom_port = fields.Integer(attribute="custom_algorithm_predict_port")
custom_evaluate_port = fields.Integer(
attribute="custom_algorithm_evaluate_port")
custom_name = fields.String(attribute="custom_algorithm_name")
custom_id_name = fields.String(attribute="custom_algorithm_alias")
custom_desc = fields.String(attribute="custom_algorithm_desc")
custom_type = fields.Function(
lambda obj: "ner"
if obj.nlp_task_id == NlpTaskEnum.extract and obj.preprocess.get(
"split_by_sentence", False) else NlpTaskEnum(obj.nlp_task_id).name)
custom_state = fields.Function(
lambda obj: StatusEnum(obj.custom_algorithm_status).name)
custom_config = fields.String(attribute="custom_algorithm_config")
created_time = fields.DateTime()
preprocess = fields.Dict()
class PredictJobSchema(Schema):
doc_type = fields.Nested(DocTypeSchema)
task_list = fields.List(fields.Nested(PredictTaskSchema))
extract_job_id = fields.Integer(attribute="predict_job_id")
extract_job_name = fields.String(attribute="predict_job_name")
extract_job_type = fields.String(attribute="predict_job_type.value")
extract_job_state = fields.Function(lambda obj: StatusEnum(obj.predict_job_status).name)
extract_job_desc = fields.String(attribute="predict_job_desc")
is_batch = fields.Boolean()
created_time = fields.DateTime()
class MarkJobSchema(Schema):
mark_job_id = fields.Integer()
mark_job_name = fields.String()
mark_job_type = fields.Function(lambda obj: obj.mark_job_type.value)
assign_mode = fields.Function(lambda obj: obj.assign_mode.value)
mark_job_state = fields.Function(
lambda obj: StatusEnum(obj.mark_job_status).name)
mark_job_desc = fields.String()
task_list = fields.List(fields.Integer())
created_time = fields.DateTime()
assessor_id = fields.Function(lambda obj: obj.reviewer_ids[0]
if len(obj.reviewer_ids) > 0 else 0)
doc_type = fields.Nested(DocTypeSchema, exclude=('doc_term_list', ))
labeler_ids = fields.List(fields.Integer(), attribute='annotator_ids')
stats = fields.Nested({
"all": fields.Integer(),
"labeled": fields.Integer(),
"audited": fields.Integer()
})
class TrainTaskSchema(Schema): # type: ignore
model_train_id = fields.Integer(attribute="train_task_id")
model_train_config = fields.Dict(attribute="train_config")
model_train_state = fields.Function(
lambda obj: StatusEnum(obj.train_status).name)
model_id = fields.Integer(attribute="train_job_id")
mark_job_ids = fields.List(fields.Integer())
train_terms = fields.List(fields.Nested(TrainTermTaskSchema))
created_time = fields.DateTime()
last_updated_time = fields.DateTime(attribute="updated_time")
model_version = fields.String()
class EvaluateTaskSchema(Schema):
model_evaluate_id = fields.Integer(attribute="evaluate_task_id")
model_evaluate_name = fields.String(attribute="evaluate_task_name")
model_evaluate_desc = fields.String(attribute="evaluate_task_desc")
model_evaluate_state = fields.Function(
lambda obj: StatusEnum(obj.evaluate_task_status).name)
model_evaluate_result = fields.Dict(attribute="evaluate_task_result")
model_id = fields.Integer(attribute="train_job_id")
mark_job_ids = fields.List(fields.Integer())
created_time = fields.DateTime()
last_updated_time = fields.DateTime(attribute="updated_time")
class TrainJobSchema(Schema): # type: ignore
model_id = fields.Integer(attribute='train_job_id')
model_name = fields.String(attribute="train_job_name")
model_desc = fields.String(attribute="train_job_desc")
status = fields.Function(lambda obj: not obj.is_deleted)
doc_type = fields.Nested(DocTypeSchema)
created_time = fields.String()
model_version = fields.String()
train_list = fields.List(fields.Nested(TrainTaskSchema))
model_evaluate = fields.Nested(EvaluateTaskSchema)
preprocess = fields.Dict()
class DocTypeSchema(Schema):
doc_terms = fields.List(fields.Integer())
doc_term_list = fields.List(fields.Nested(DocTermSchema))
doc_relation_list = fields.List(fields.Nested(EntityDocRelationSchema))
doc_lexicon_list = fields.List(fields.Nested(WordsegDocLexiconSchema),
attribute='doc_rules')
doc_type_id = fields.Integer()
doc_type_name = fields.String()
doc_type_desc = fields.String()
is_top = fields.Boolean(attribute="is_favorite")
created_time = fields.DateTime()
group_id = fields.Integer()
status = fields.Function(lambda obj: not obj.is_deleted)
class DocSchema(Schema): # type: ignore
doc_id = fields.Integer()
doc_unique_name = fields.String()
doc_raw_name = fields.String()
convert_state = fields.Function(
lambda obj: StatusEnum(obj.doc_status).name)