Fix NaN and "None" string serialization in table data API respons

.gitignore

@@ -283,3 +283,5 @@ backend/target/media/
 pg_data/
 data/
 log/
+
+docker-compose.yml

backend/target/metadata/api/views.py

@@ -1,3 +1,5 @@
+import math
+
 from django.conf import settings
 from rest_framework import status
 from rest_framework.decorators import action
@@ -12,7 +14,6 @@
 
 from .serializers import ColumnSerializer
 from .serializers import NestedTableSerializer
-from .serializers import ResumedColumnSerializer
 from .serializers import SchemaSerializer
 from .serializers import SettingsSerializer
 from .serializers import TableSerializer
@@ -21,8 +22,6 @@
 class TableRegistrationError(Exception):
     """Raised when table registration fails"""
 
-    pass
-
 
 class TableDeletePermissionError(PermissionError):
     """Raised when a user tries to delete a table without permission"""
@@ -149,7 +148,8 @@ def register_table(self, user, data):
             tablename=data.get("name"),
         ):
             table_name = f"{data.get('schema')}.{data.get('name')}"
-            raise TableRegistrationError(f"Table {table_name} not found in database")
+            msg = f"Table {table_name} not found in database"
+            raise TableRegistrationError(msg)
 
         # Tamanho da tabela e quantidade de linhas estimadas.
         stats = db.get_table_status(
@@ -206,7 +206,7 @@ def register(self, user, data):
             if table.catalog_type == Table.CATALOG_TYPE_CLUSTER:
                 related_tablename = data.get("related_table_name", None)
                 if not related_tablename:
-                    raise MissingRelatedTableError()
+                    raise MissingRelatedTableError  # noqa: TRY301
 
                 # region Register related table if not registered
                 if related_tablename:
@@ -264,7 +264,7 @@ def create(self, request):
             data = self.get_serializer(instance=table).data
             return Response(data, status=status.HTTP_201_CREATED)
 
-        except Exception as e:
+        except Exception as e:  # noqa: BLE001
             content = {"error": str(e)}
             return Response(content, status=status.HTTP_500_INTERNAL_SERVER_ERROR)
 
@@ -356,7 +356,9 @@ def pending_registration(self, request):
         return Response({}, status=status.HTTP_200_OK)
 
     def check_mandatory_ucds(self, table_ucds, required_ucds):
-        """Verifica se todos os UCDs obrigatórios estão presentes e têm colunas válidas."""
+        """Verifica se todos os UCDs
+        obrigatórios estão presentes e têm colunas válidas.
+        """
         missing = []
         for ucd in required_ucds:
             column = table_ucds.get(ucd)
@@ -379,7 +381,8 @@ def complete_registration(self, request, pk=None):
             # Check if related table have all required UCDs assigned
             related_ucds = self.get_table_ucds(table.related_table)
             missing = self.check_mandatory_ucds(
-                related_ucds, Table.RELATED_REQUIRED_UCDS
+                related_ucds,
+                Table.RELATED_REQUIRED_UCDS,
             )
             if len(missing) > 0:
                 return Response(
@@ -430,42 +433,46 @@ def parse_filters(self, query_params):
 
     def perform_destroy(self, instance):
         if instance.schema.owner != self.request.user:
-            raise TableDeletePermissionError()
+            raise TableDeletePermissionError
 
         if instance.related_table:
             instance.related_table.delete()
 
         instance.delete()
         return Response(status=status.HTTP_204_NO_CONTENT)
 
-    def convert_bigints_to_string(self, data, bigint_columns):
+    def sanitize_rows(self, data, bigint_columns=None):
+        if isinstance(data, list):
+            return [self.sanitize_rows(item, bigint_columns) for item in data]
         if isinstance(data, dict):
-            return {
-                k: str(v)
-                if k in bigint_columns and isinstance(v, int) and v > 9007199254740991
-                else v
-                for k, v in data.items()
-            }
-        elif isinstance(data, list):
-            return [
-                self.convert_bigints_to_string(item, bigint_columns) for item in data
-            ]
+            result = {}
+            for k, v in data.items():
+                if (isinstance(v, float) and not math.isfinite(v)) or (
+                    isinstance(v, str) and v == "None"
+                ):
+                    result[k] = None
+                elif (
+                    bigint_columns
+                    and k in bigint_columns
+                    and isinstance(v, int)
+                    and v > 9007199254740991  # noqa: PLR2004
+                ):
+                    result[k] = str(v)
+                else:
+                    result[k] = self.sanitize_rows(v, bigint_columns)
+            return result
         return data
 
     @action(detail=True, methods=["get"])
     def data(self, request, pk=None):
-        # print("-----------------------------")
-
         # IMPORTANTE: Não pode ser utilizado o self.get_object()
         # por que falha se um dos campos de filtro for "id"
         # pk é a identificação que vem na url /{pk}/data/
         # e não é afetada pelos filtros.
         queryset = self.get_queryset()
         table = queryset.prefetch_related("columns").get(pk=pk)
 
-        # print(table)
         ucds = self.get_table_ucds(table)
-        # print(ucds)
 
         # Total de linhas estimado da tabela.
         count = table.nrows
@@ -476,22 +483,14 @@ def data(self, request, pk=None):
             "pageSize",
             int(settings.REST_FRAMEWORK["PAGE_SIZE"]),
         )
-        # print("Page: ", page)
-        # print("Page Size: ", page_size)
 
         limit = int(page_size)
         offset = (limit * page) - limit
 
-        # print("Offset: ", offset)
-        # print("Limit: ", limit)
-
-        # # TODO: selecionar as colunas que serao utilizadas.
-        # columns = request.query_params.get("columns", "")
-        # columns = columns.split(",")
+        # TODO: selecionar as colunas que serao utilizadas.
 
         # Parse Filters
         url_filters = self.parse_filters(request.query_params)
-        # print("Filters: ", url_filters)
 
         ordering = request.query_params.get("ordering", None)
 
@@ -509,9 +508,15 @@ def data(self, request, pk=None):
             row.update(
                 {
                     "meta_catalog_id": table.id,
-                    "meta_id": str(row.get(ucds.get("meta.id;meta.main"))),
-                    "meta_ra": str(row.get(ucds.get("pos.eq.ra;meta.main"))),
-                    "meta_dec": str(row.get(ucds.get("pos.eq.dec;meta.main"))),
+                    "meta_id": None
+                    if (v := row.get(ucds.get("meta.id;meta.main"))) is None
+                    else str(v),
+                    "meta_ra": None
+                    if (v := row.get(ucds.get("pos.eq.ra;meta.main"))) is None
+                    else str(v),
+                    "meta_dec": None
+                    if (v := row.get(ucds.get("pos.eq.dec;meta.main"))) is None
+                    else str(v),
                     "meta_radius_arcmin": row.get(ucds.get("phys.angSize;src")),
                 },
             )
@@ -523,17 +528,11 @@ def data(self, request, pk=None):
             if datatype in ["bigint", "int8"]:
                 bigint_columns.append(col.name)
 
-        parsed_rows = self.convert_bigints_to_string(rows, bigint_columns)
-
-        # columns_data = ResumedColumnSerializer(
-        #     table.columns.all(),
-        #     many=True,
-        # ).data
+        parsed_rows = self.sanitize_rows(rows, bigint_columns)
 
         results = {
             "results": parsed_rows,
             "count": count,
-            # "columns": columns_data,
             "has_more": (offset + limit) < count,
         }