kodexa 7.4.413092627931__py3-none-any.whl → 7.4.413159458494__py3-none-any.whl

kodexa/dataclasses/__init__.py

@@ -54,21 +54,23 @@ class LLMDataAttribute(BaseModel):
         self.node_uuid_list = source.node_uuid_list
         self.page_number = source.page_number
 
-    def create_exception(
-            self,
-            exception_type_id: str,
-            exception_type: str,
-            normalized_text: str,
-            message: str,
-            exception_detail: str,
-    ):
-        content_exception = ContentException(
-            exception_type=exception_type,
-            exception_detail=exception_detail,
-            message=message,
-            tag_uuid=self.tag_uuid,
-        )
-        self.exceptions.append(content_exception)
+    def process_exceptions(self, document: "KodexaDocumentLLMWrapper"):
+        # Lets make sure we add all the content exceptions
+        if self.exceptions is not None:
+            for exception in self.exceptions:
+                # We have two types of exception, one in the API and one in the
+                # document
+                from kodexa.model import ContentException as KodexaContentException
+                internal_exception = KodexaContentException(
+                    tag=exception.tag,
+                    exception_type=exception.exception_type,
+                    message=exception.message,
+                    exception_details=exception.exception_details,
+                    severity=exception.severity,
+                    group_uuid=self.group_uuid,
+                    tag_uuid=self.tag_uuid,
+                )
+                document.doc.add_exception(internal_exception)
 
     def to_dict(self, taxonomy: Taxonomy) -> dict:
         """Convert attribute to JSON with normalized value"""
@@ -111,6 +113,24 @@ class LLMDataObject(BaseModel):
     class Config:
         arbitrary_types_allowed = True
 
+    def process_exceptions(self, document: "KodexaDocumentLLMWrapper"):
+        # Lets make sure we add all the content exceptions
+        if self.exceptions is not None:
+            for exception in self.exceptions:
+                # We have two types of exception, one in the API and one in the
+                # document
+                from kodexa.model import ContentException as KodexaContentException
+                internal_exception = KodexaContentException(
+                    tag=exception.tag,
+                    exception_type=exception.exception_type,
+                    message=exception.message,
+                    exception_details=exception.exception_details,
+                    severity=exception.severity,
+                    group_uuid=self.group_uuid,
+                    tag_uuid=self.tag_uuid,
+                )
+                document.doc.add_exception(internal_exception)
+
     def get_all_review_pages(self):
         """
         Returns a list of unique page numbers that would be included in the review.
@@ -118,7 +138,7 @@ class LLMDataObject(BaseModel):
         :return: list of unique page numbers
         """
         pages = set()
-        for field in self.__fields__:
+        for field in self.model_fields:
             pages.update(self._get_field_pages(field))
         return sorted(list(pages))
 
@@ -169,7 +189,7 @@ class LLMDataObject(BaseModel):
     def to_dict(self, taxonomy: Taxonomy) -> dict:
         """Convert data object to JSON using normalized values and taxon paths"""
         result = {}
-        for field in self.__fields__:
+        for field in self.model_fields:
             value = getattr(self, field)
 
             if isinstance(value, list) and len(value) > 0:
@@ -177,7 +197,7 @@ class LLMDataObject(BaseModel):
                     # We need to find the first field of the object that is a LLMDataAttribute
                     # and use that to derive the taxon path of the LLMDataObject
                     data_attribute = None
-                    for child_field in value[0].__fields__:
+                    for child_field in value[0].model_fields:
                         child_attr = getattr(value[0], child_field)
                         if isinstance(child_attr, LLMDataAttribute):
                             data_attribute = child_attr
@@ -205,7 +225,7 @@ class LLMDataObject(BaseModel):
         :return: dict of this data object and children for the specified page
         """
         review = {}
-        for field in self.__fields__:
+        for field in self.model_fields:
             review_field = self._build_review(field, page_number)
             if review_field:
                 review[field] = review_field
@@ -234,26 +254,6 @@ class LLMDataObject(BaseModel):
 
         return None
 
-    def create_exception(
-            self,
-            exception_type_id: str,
-            exception_type: str,
-            message: str,
-            exception_detail: str,
-            severity: str = "ERROR",
-    ):
-        content_exception = ContentException(
-            exception_type=exception_type,
-            exception_details=exception_detail,
-            message=message,
-            group_uuid=self.group_uuid,
-            severity=severity,
-        )
-        if self.exceptions is None:
-            self.exceptions = []
-
-        self.exceptions.append(content_exception)
-
     def apply_labels(
             self, document: "KodexaDocumentLLMWrapper", parent_group_uuid: str = None,
             assistant: Optional["Assistant"] = None
@@ -275,24 +275,11 @@ class LLMDataObject(BaseModel):
         """
 
         # Lets make sure we add all the content exceptions
-        if self.exceptions is not None:
-            for exception in self.exceptions:
-                # We have two types of exception, one in the API and one in the
-                # document
-                from kodexa.model import ContentException as KodexaContentException
-                internal_exception = KodexaContentException(
-                    exception_type=exception.exception_type,
-                    message=exception.message,
-                    exception_details=exception.exception_details,
-                    severity=exception.severity,
-                    group_uuid=exception.group_uuid,
-                    tag_uuid=exception.tag_uuid,
-                )
-                document.doc.add_exception(internal_exception)
+        self.process_exceptions(document)
 
         # Let's go through this data object and find all the attributes that have a value
         # then we will apply the labels to the document
-        for field in self.__fields__:
+        for field in self.model_fields:
             logger.info(f"Processing field {field}")
             value = getattr(self, field)
 
@@ -311,8 +298,6 @@ class LLMDataObject(BaseModel):
             # We need to add the label to the document for this attribute
 
             tag = value.taxon_path
-
-            # TODO need to work out why we are missing them?
             logger.info(f"Value: {value.normalized_text}, node_uuid_list: {value.node_uuid_list}")
             if value.node_uuid_list is None:
                 value.node_uuid_list = value.line_ids
@@ -361,31 +346,16 @@ class LLMDataObject(BaseModel):
                         current_value.append(new_tag)
                         node.remove_feature("tag", tag)
                         node.add_feature("tag", tag, current_value, single=False)
-                        # try:
-                        #     if value.data_type == 'Derived':
-                        #         logger.info(f"Node already has tag {tag} - Tagging something nearby {node.get_all_content()}")
-                        #         nearby_node = find_nearby_word_to_tag(node, tag)
-                        #         nearby_node.tag(
-                        #             tag_to_apply=tag,
-                        #             value=value.normalized_text,
-                        #             tag_uuid=tag_uuid,
-                        #             cell_index=self.cell_index,
-                        #             selector="//word",
-                        #             confidence=-1,
-                        #             group_uuid=self.group_uuid,
-                        #             parent_group_uuid=parent_group_uuid,
-                        #             owner_uri=f"assistant://{assistant.id}" if assistant else f"model://taxonomy-llm",
-                        #         )
-                        #     else:
-                        #         logger.info(f"Node already has tag {tag} - Skipping.")
-                        # except:
-                        #     logger.error(f"Error tagging nearby node with tag {tag}")
 
             logger.info(f"Applied label {tag} to {len(nodes_to_label)} nodes")
+
+            # Lets make sure we add all the content exceptions
+            self.process_exceptions(document)
+
         if isinstance(value, LLMDataObject):
             # We need to apply the labels to the document for this object
             value.apply_labels(document, parent_group_uuid=self.group_uuid)
-            # logger.info(f"Applied labels to data object {value.group_uuid}")
+            logger.info(f"Applied labels to data object {value.group_uuid}")
 
 
 def find_nearby_word_to_tag(node, tag):

{kodexa-7.4.413092627931.dist-info → kodexa-7.4.413159458494.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: kodexa
-Version: 7.4.413092627931
+Version: 7.4.413159458494
 Summary: Python SDK for the Kodexa Platform
 Author: Austin Redenbaugh
 Author-email: austin@kodexa.com

{kodexa-7.4.413092627931.dist-info → kodexa-7.4.413159458494.dist-info}/RECORD

@@ -3,7 +3,7 @@ kodexa/assistant/__init__.py,sha256=nlXm_YnV_50hgn0TIT2Fkc2fQ-86OjmctY_j8My9nc4,
 kodexa/assistant/assistant.py,sha256=5KFdbqFSLIZJyDRyZdpcfr448fT-CW4JhYu9A6B9DGY,14663
 kodexa/connectors/__init__.py,sha256=WF6G_MUeU32TlKSUKkpNoNX7dq8iBPliFMep4E8BmZc,328
 kodexa/connectors/connectors.py,sha256=FpUZDkSyHld2b9eYRuVOWzaFtuGoaRuPXXicJB7THbc,10413
-kodexa/dataclasses/__init__.py,sha256=l1bpG3wyGEwq9SpdD9uVFuIep1A9x0si1eqxCmyWH68,20641
+kodexa/dataclasses/__init__.py,sha256=CHMNsOamWA3gY5203gn8Ef5q1fgcczMtWKEvNjIOzPs,19486
 kodexa/dataclasses/templates/llm_data_class.j2,sha256=YWjStW136chV_59JM3AYis3i-0jdrqDvLXsISUW9zDU,660
 kodexa/model/__init__.py,sha256=rtLXYJBxB-rnukhslN9rlqoB3--1H3253HyHGbD_Gc8,796
 kodexa/model/base.py,sha256=CaZK8nMhT1LdCpt4aLhebJGcorjq9qRID1FjnXnP14M,521
@@ -44,7 +44,7 @@ kodexa/testing/test_utils.py,sha256=v44p__gE7ia67W7WeHN2HBFCWSCUrCZt7G4xBNCmwf8,
 kodexa/training/__init__.py,sha256=xs2L62YpRkIRfslQwtQZ5Yxjhm7sLzX2TrVX6EuBnZQ,52
 kodexa/training/train_utils.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 kodexa/utils/__init__.py,sha256=Pnim1o9_db5YEnNvDTxpM7HG-qTlL6n8JwFwOafU9wo,5928
-kodexa-7.4.413092627931.dist-info/LICENSE,sha256=WNHhf_5RCaeuKWyq_K39vmp9F28LxKsB4SpomwSZ2L0,11357
-kodexa-7.4.413092627931.dist-info/METADATA,sha256=GFaqjOkldJymEr9odh1g5d0XbTcrx91KocpbuWPcqO0,3528
-kodexa-7.4.413092627931.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
-kodexa-7.4.413092627931.dist-info/RECORD,,
+kodexa-7.4.413159458494.dist-info/LICENSE,sha256=WNHhf_5RCaeuKWyq_K39vmp9F28LxKsB4SpomwSZ2L0,11357
+kodexa-7.4.413159458494.dist-info/METADATA,sha256=cGd2yP7RRGfQviKkswuRnvAKwyvcGX-tZ_tb_kEN-5c,3528
+kodexa-7.4.413159458494.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
+kodexa-7.4.413159458494.dist-info/RECORD,,