ApiLogicServer 14.3.14__py3-none-any.whl → 14.3.25__py3-none-any.whl

api_logic_server_cli/prototypes/manager/system/genai/examples/genai_demo/genai_demo_docs_logic/api/system/custom_endpoint.py

@@ -0,0 +1,1085 @@
+from __future__ import annotations  # enables Resource self reference
+import sqlalchemy
+import logging
+import contextlib
+from sqlalchemy import Column, Table, ForeignKey
+from sqlalchemy.orm.decl_api import DeclarativeMeta #sqlalchemy.orm.decl_api.DeclarativeMeta
+from sqlalchemy.orm import relationships, relationship
+from sqlalchemy.ext.declarative import declarative_base
+from sqlalchemy_utils import get_referencing_foreign_keys
+from sqlalchemy import event, MetaData, and_, or_
+from sqlalchemy.inspection import inspect
+from sqlalchemy.sql import text
+from flask import jsonify
+from sqlalchemy_utils.query_chain import QueryChain
+import flask_sqlalchemy
+import safrs
+from safrs.errors import JsonapiError, ValidationError
+from security.system.authorization import Security
+from typing import List, Dict, Tuple
+import json 
+import requests
+import config.config as config
+from config.config import Args
+from api.system.expression_parser import parsePayload
+
+resource_logger = logging.getLogger("api.customize_api")
+
+db = safrs.DB 
+"""this is a safrs db not DB"""
+
+session = db.session  # type: sqlalchemy.orm.scoping.scoped_session
+
+
+class DotDict(dict):
+    """ dot.notation access to dictionary attributes """
+    # thanks: https://stackoverflow.com/questions/2352181/how-to-use-a-dot-to-access-members-of-dictionary/28463329
+    __getattr__ = dict.get
+    __setattr__ = dict.__setitem__
+    __delattr__ = dict.__delitem__
+
+
+class CustomEndpoint():
+    """
+    Nested CustomEndpoint Definition.  Internal system services for Ontimize.
+
+        customer = CustomEndpoint(model_class=models.Customer, alias="Customer"
+        , fields = [(models.Customer.CompanyName, "Customer Name")] 
+        , children = [
+            CustomEndpoint(model_class=models.Order, alias = "orders"
+                , join_on=models.Order.CustomerId
+                , fields = [(models.Order.AmountTotal, "Total"),(models.Order.ShippedDate, "Ship Date")]
+                , children = CustomEndpoint(model_class=models.OrderDetail, alias="details"
+                    , join_on=models.OrderDetail.OrderId
+                    , fields = [models.OrderDetail.Quantity, models.OrderDetail.Amount]
+                    , children = CustomEndpoint(model_class=models.Product, alias="data"
+                        , join_on=models.OrderDetail.ProductId
+                        , fields=[models.Product.UnitPrice, models.Product.UnitsInStock]
+                        , isParent=True
+                        , isCombined=False
+                    )
+                )
+            ),
+            CustomEndpoint(model_class=models.OrderAudit, alias="orderAudit")  # sibling child
+            ]
+        )
+        result = customer.execute(customer,"", "ALFKI")
+        # or
+        #result = customer.get(request,"OrderList&OrderList.OrderDetailList&OrderList.OrderDetailList.Product", "ALFKI")
+    """
+
+    def __init__(self
+            , model_class: DeclarativeMeta | None
+            , alias: str = ""
+            , fields: list[tuple[Column, str] | Column] = []
+            , children: list[CustomEndpoint] | CustomEndpoint = []
+            , join_on: list[tuple[Column] | Column] = None
+            , calling: callable = None
+            , filter_by: str = None
+            , order_by: Column = None
+            , isParent: bool = False
+            , isCombined: bool = False
+            , pagesize: int = 100
+            , offset: int = 0
+            ):
+        """
+
+        Declare a custom user shaped resource.
+
+        Args:
+            :model_class (DeclarativeMeta | None): model.TableName
+            :alias (str, optional): _description_. Defaults to "".
+            :fields (list[tuple[Column, str]  |  Column], optional): model.Table.Column. Defaults to [].
+            :children (list[CustomEndpoint] | CustomEndpoint, optional): CustomEndpoint(). Defaults to []. (OneToMany)
+            join_on: list[tuple[Column, Column]] - this is a tuple of parent/child multiple field joins 
+            :calling - name of function (passing row for virtual attributes or modification)
+            :filter_by is string object in SQL format (e.g. '"ShipDate" != null')
+            :order_by is Column object used to sort aac result (e.g. order_by=models.Customer.Name)
+            :isParent = if True - use parent foreign key to join single lookup (ManyToOne)
+            :isCombined =  combine the fields of the isParent = routeTrue with the _parentResource (flatten) 
+            
+        """
+        if not model_class:
+            raise ValueError("CustomEndpoint model_class=models.EntityName is required")
+        #if join_on is None and children is not None or parent is not None:
+        #    raise ValueError("join_on= is required if using children (child column)")
+
+        self._model_class = model_class
+        self.alias = alias or model_class._s_type.lower()
+        self.fields = fields
+        self.children = children or []
+        self.calling = calling
+        self.filter_by = filter_by
+        self.order_by = order_by
+        self.isCombined = isCombined
+        self.join_on = join_on 
+        self.isParent= isParent 
+        self.pagesize = pagesize
+        self.offset = offset
+        self.totalQueryRecordsNumber =  999
+        self.startRecordIndex = 0
+        if isinstance(join_on, tuple):
+            if len(join_on) > 0:
+                # get parent or child
+                self.foreignKey = join_on[0] # - if we have multiple joins
+        else:
+            self.foreignKey = join_on
+
+        # Internal Private 
+        self.primaryKey: str = inspect(model_class).primary_key[0].name
+        self.primaryKeyType: str = inspect(self._model_class).primary_key[0].type
+        self._model_class_name: str = self._model_class._s_class_name
+        # inspect(model_class).primary_key[0].type = Integer() ...
+        self._parentResource: CustomEndpoint = None  # ROOT
+        self._pkeyList: list = [] # primary key list  collect when needed - do not store
+        self._fkeyList: list = [] # foreign_key list (used by isParent)  collect when needed - do not store
+        self._dictRows: list = [] # temporary holding for query results (Phase 1)
+        self._parentRow = Dict[str, any] # keep track of linkage
+        self._method = None
+        self._href = None
+        self._columnNames = [k.key for k in self._model_class._s_columns]
+        key = self._model_class_name
+        from api.api_discovery.ontimize_api import getMetaData
+        resources = getMetaData(key)
+        self._attributes = resources["resources"][key]["attributes"]
+        self._quote = '`' if Args.backtic_as_quote else '"'
+        
+    def __str__(self):
+            return  f"Alias {self.alias} Model: {self._model_class.__name__} PrimaryKey: {self.primaryKey} FilterBy: {self.filter_by} OrderBy: {self.order_by}"
+
+    def get(self: CustomEndpoint, request: safrs.request.SAFRSRequest, include: str, altKey: str = None) -> dict:
+        """_summary_
+
+        Args:
+            :self (CustomEndpoint): 
+            :request (safrs.request.SAFRSRequest): 
+            :include (str): name(s) of relationship from swagger include=
+            :altKey (str, optional):  Defaults to None.
+
+        Returns:
+            dict: JSON result
+        """
+        if request.method == 'OPTIONS':
+            return jsonify(success=True)
+        
+        serverURL = f"{request.host_url}api"
+        query = f"{serverURL}/{self._model_class_name}?include={include}"
+        args = request.args
+        key, value,  limit, offset, order_by, filter_ = self.parseArgs(args)
+        if altKey is not None:
+            query += f"&filter%5B{self.primaryKey}%5D={altKey}"
+        elif key is not None and value is not None:
+            query += f"&filter%5B{key}%5D={value}"
+        else: 
+            #query =  query if filter_ is None or filter_ is '1=1' else f"{query} and {filter_}"
+            query = (
+                query
+                if filter_ is None or filter_ == '1=1'
+                else f"{query} and {filter_}"
+            )
+        self._href = f"{request.url_root[:-1]}{request.path}"
+        print(f"limit: {limit}, offset: {offset}, sort: {order_by}, query: {query}")
+        params = {'page[limit]': limit, 'page[offset]': offset}
+        resource_logger.debug(f"CustomEndpoint get using query: {query}")
+        if Args.security_enabled:
+            jwt = request.headers.get("Authorization") or ""
+            header = {"Authorization": jwt,"Content-Type": "application/json"}
+            result = requests.get(query, headers=header, params=params)
+        else:
+            result = requests.get(query, params=params)
+        if result.status_code == 200:
+            jsonResult = json.loads(result.content)
+            self._populateResponse(jsonResult) # Pass the JSON result to CustomEndpoint 
+            result = self.execute(request) # a dict of args
+            return result
+        return {"error": result.status_code}
+        
+    def execute(self: CustomEndpoint, request: safrs.request.SAFRSRequest, altKey: str = None) -> dict:
+        """
+        execute a model_class resource 
+        Args:
+            :self (CustomEndpoint): 
+            : request SarfsRequest (holds args and data payload, etc)
+            : altKey optional <key> passed from @app.route("/foo/<altKey>")
+            
+            'page[offset]=0 or offset=0
+            'page[limit]=10 or limit=10
+            'sort=CompanyName'
+            'filter[id]=ALFKI or Id=ALFKI
+            
+        #self._model_class.get_instance("ALFKI") returns the Cusomter/OrderList/OrderDetailList and Product
+        #util.row_to_dict(self._model_class.get_instance("ALFKI").OrderList[0].OrderDetailList[0].Product)
+        #for f in util.row_to_dict(self._model_class.get_instance("ALFKI")).get("relationships") :print(f'{util.row_to_dict(self._model_class.get_instance("ALFKI"))}.{f}')
+        #self._model_class.get_instance('ALFKI')._s_relationships.get("OrderList").synchronize_pairs
+        #"Manager" in self._model_class._s_relationships.keys() #_s_jsonapi_attrs.keys()
+        #rom .jsonapi_formatting import jsonapi_filter_query, jsonapi_filter_list, jsonapi_sort, jsonapi_format_response, paginate
+        #self._model_class.__mapper__.relationships.get("Manages").primaryjoin.left or right left.key or right.key
+    
+        '''
+
+        Returns:
+            dict: data dict from sql
+        """
+        expressions = []
+        args = {}
+        payload = {}
+        result = {}
+        jwt = ""
+        pkey = None
+        value = None
+        if request is not None:
+            jwt = request.headers.get("Authorization") or ""
+            method = request.method
+            self._method = method
+            args = request.args
+            if len(request.data) > 0:
+                payload = json.loads(request.data.decode('utf-8'))
+            self._printIncludes(1) # debug print
+            if method == 'DELETE':
+                raise ValidationError( 'Delete is not supported at this time')
+            elif method == 'OPTIONS':
+                return jsonify(success=True)
+            elif method in ["PUT","PATCH"]:
+                api = "api" # TODO Args.api_prefix()
+                serverURL = f"{request.host_url}{api}"
+                url = f"{serverURL}/{self._model_class_name}"
+                return self.handlePayload(method=method, payload=payload, url=url, jwt=jwt, altKey=altKey)
+            elif method == 'POST':
+                try:
+                    return self.insert_or_update(payload=payload, altKey=altKey)
+                except Exception as ex:
+                    raise ValidationError( f'{method} error on entity {self._model_class_name} msg: {ex}') from ex
+            elif method == 'GET':
+                if payload:
+                    expressions, filter_, columns, sqltypes, offset, limit, order_by, data = parsePayload(clz=self._model_class, payload=payload)
+                else:
+                    pkey , value,  limit, offset, order_by , filter_  = self.parseArgs(args)
+        #serverURL = f"{request.host_url}api"
+        #query = f"{serverURL}/{self._model_class_name}"
+        self.startRecordIndex = int(offset)
+        resource_logger.debug(f"CustomEndpoint execute on: {self._model_class_name} using alias: {self.alias}")
+        filter_by = None
+        #key = args.get(pkey) if args.get(pkey) is not None else args.get(f"filter[{pkey}]")
+        _quote = '`' if Args.backtic_as_quote else '"'
+        if value is not None and value != 'undefined':
+            filter_by = f'{_quote}{pkey}{_quote} = {self.quoteStr(value)}'
+            self._pkeyList.append(self.quoteStr(value))
+        elif altKey is not None:
+            filter_by = f'{_quote}{pkey}{_quote} = {self.quoteStr(altKey)}'
+            self._pkeyList.append(self.quoteStr(altKey))
+        filter_by = filter_by if filter_ is None else f"{filter_by} and {filter_}" if filter_by is not None else filter_
+        self._href = f"{request.url_root[:-1]}{request.path}"
+        limit =  max(self.pagesize, int(limit))
+        print(f"limit: {limit}, offset: {offset}, sort: {order_by},filter_by: {filter_by}, add_filter {filter_}")
+        try:
+            self._createRows(limit=limit,offset=offset,order_by=order_by,filter_by=filter_by, expressions=expressions) 
+            self._executeChildren()
+            self._modifyRows(result)
+            return json.dumps(result, indent=4, ensure_ascii=False).encode('utf8')
+        except Exception as ex:
+            resource_logger.error(f"CustomEndpoint error {ex}")
+            return f"'error': {ex}"
+
+    def _executeChildren(self):
+        """
+        Recursive execution of included CustomEndpoints
+        """
+        if isinstance(self.children, CustomEndpoint):
+            self.children._parentResource = self
+            self.children._href = f"{self.modifyPath(self._href)}{self.children._model_class_name}"
+            self.children._processChildren()
+        elif len(self.children) > 0:
+            for child in self.children:
+                child._parentResource = self
+                child._href = f"{self.modifyPath(self._href)}{child._model_class_name}"
+                child._processChildren()
+        
+    def _collectPKeys(self, keyName)-> list:
+        keyList = []
+        if keyName is None:
+            return keyList
+        for row in self._dictRows:
+            if keyName in row:
+                key = row.get(keyName)
+                if key is not None and key not in keyList:
+                    keyList.append(key)
+
+        self._pkeyList = keyList
+        return keyList
+        
+    def _collectParentKeys(self, keyName: str) -> list:
+        keyList = []
+        #if not self.isParent or keyName is None:
+        if self._parentResource is None or keyName is None:
+            return self._parentResource._pkeyList
+        for row in self._parentResource._dictRows:
+            if keyName in row:
+                key = row.get(keyName)
+                if key is not None and key not in keyList:
+                    keyList.append(key)
+
+        self._fkeyList = keyList
+        return keyList
+
+    def _createRows(self,limit:int = 10, offset = 0, filter_by: str = None, order_by: str = None, expressions: list = []):
+        """
+        execute and store rows based on list of keys in model
+        :limit = 10
+        :offset = 0
+        :filter_by root only
+        :order_by root only
+        :expressions = list of expressions
+        """
+        # If _populateResponse is used - the _dictRows are already filled
+        # or the parent resource has now rows - so no need to fetch
+        if len(self._dictRows) > 0 or \
+            (self._parentResource is not None \
+            and len(self._parentResource._dictRows) == 0):
+            return
+        model_class = self._model_class
+        model_class_name = self._model_class_name
+        queryFilter = self._createFilterFromKeys()
+        session_qry= session.query(model_class)
+        #result = session.execute(select(model_class).where(text("Id = 'ALFKI'"))).all() #.join(models.Customer.OrderList)).order()
+        if queryFilter is None or queryFilter == 'None':
+            #query = select(self._model_class)
+            resource_logger.debug(
+                    f"CreateRows on {model_class_name} using filter_by: {self.filter_by} order_by: {self.order_by}")
+            if self.filter_by is not None:
+                qry = session_qry.filter(text(self.filter_by))
+                if self.order_by is not None:
+                    qry = qry.order_by(self.order_by)
+                if filter_by is not None and 'undefined' not in filter_by:
+                    resource_logger.debug(
+                    f"Adding filter_by: {filter_by}")
+                    qry = qry.filter(text(filter_by))
+                rows = qry.limit(limit).offset(offset).all()
+            else:
+                if filter_by is not None:
+                    resource_logger.debug(
+                    f"Adding filter_by: {filter_by}")
+                    session_qry = session_qry.filter(text(filter_by))
+                
+                if order_by:
+                    if isinstance(order_by, list) and len(order_by) > 0:
+                        col_name = order_by[0]["columnName"]
+                        direction =  order_by[0]["ascendent"] if "ascendent" in order_by[0] else False
+                        for a in self._attributes:
+                            if a['attr'].key == col_name:
+                                col_name = f"{a['attr'].columns[0].name } { 'asc' if direction else 'desc'}"
+                                q = '"'
+                                col_name = f"{q}{a['attr'].columns[0].name}{q}"
+                                break
+                        session_qry = session_qry.order_by(text(col_name))
+                    else:
+                        if order_by in self._attributes:
+                            session_qry = session_qry.order_by(text(order_by))
+                rows = session_qry.limit(limit).offset(offset).all()
+        else:
+            resource_logger.debug(
+                f"CreateRows on {model_class_name} using QueryFilter: {queryFilter} order_by: {self.order_by}")
+            if self.order_by is not None:
+                    if self.order_by in self._attributes:
+                        session_qry = session_qry.filter(text(queryFilter)).order_by(self.order_by)
+            elif  self.filter_by is None:
+                session_qry = session_qry.filter(text(queryFilter))
+            else:
+                if filter_by:
+                    resource_logger.debug(
+                    f"Adding on {model_class_name} using filter_by: {filter_by}")
+                    session_qry = session_qry.filter(text(filter_by))#.filter(text(self.filter_by))
+            if order_by:
+                col_name = order_by[0]["columnName"]
+                for a in self._attributes:
+                    if a['attr'].key == col_name:
+                        col_name = a['attr'].columns[0].name
+                        break
+                session_qry = session_qry.order_by(text(col_name))
+            rows = session_qry.limit(limit).offset(offset).all()
+        if rows:    
+            dictRows = self.rows_to_dict(rows)
+            self._dictRows = dictRows
+        
+
+    def _createFilterFromKeys(self):
+        aFilter = None
+        if self.join_on:
+            """
+            join_on=[(models.SourceDatum.clientId, models.SourceDatum.clientId),(models.SourceDatum.dataYear, models.SourceDatum.priorYear)]
+            we may have multiple joins - need to collect each one 
+            #clientId in (clientId keys) and  dataYear in (priorYear keys)
+            #result += F"{and} " + self._extractedFromKeys(keyName, keys)
+            
+        """     
+        if isinstance(self.join_on, list):
+            andOp = ""
+            for join in self.join_on:
+                aFilter = self.buildJoin(andOp, join)
+                andOp = " and "
+        else:
+            aFilter = self.buildJoin("", self.join_on)        
+    
+        return aFilter
+
+    def buildJoin(self, andOp: str, join: Column) -> str:
+        joinStr = None
+        if join is not None:
+            if join.__class__.__name__ == 'InstrumentedAttribute':
+                if hasattr(join,"prop") and join.prop.__class__.__name__ == 'RelationshipProperty':
+                #    pass #oin.prop._join_condition.foreign_key_columns
+                    for l in join.prop._join_condition.foreign_key_columns: 
+                        fkeyName = self.primaryKey if self.isParent else l.key
+                        self.foreignKey = l
+                        keyName = l.key if self.isParent else self.primaryKey
+                else:
+                    fkeyName = self.primaryKey if self.isParent else join.key #child - parent pkey is implied
+                    keyName = join.key if self.isParent else self.primaryKey
+                
+            elif len(join) == 2:
+                pkeyName = join[0].key #parent
+                fkeyName = join[1].key #child
+                self.primaryKey = fkeyName if self.isParent else self.primaryKey
+                self.foreignKey = join[1] if self.isParent else join[0]
+                keyName = join[1].key if self.isParent else pkeyName
+            
+            keys = self._collectParentKeys(keyName)
+            if keys is not None:
+                joinStrKeys = self._extractedFromKeys(fkeyName , keys)
+            return f"{andOp}{joinStrKeys}"
+        return None
+            
+
+    def _extractedFromKeys(self, keyName: str, keys: object):
+        if keys is None or len(keys) == 0:
+            return None
+        keys = f"{keys}"
+        keys = keys.replace("[", "")
+        keys = keys.replace("]", "")
+        result = ""
+        if len(keys) > 0:
+            result = f'{keyName} in ({keys})'
+            result = result.replace(".", "\".\"")
+        if self.filter_by is not None:
+            result += f" and text({self.filter_by})"
+        return result
+
+    def _printIncludes(self, level: int):
+        parenName = self._parentResource._model_class_name if self._parentResource is not None else "None"
+        if self.foreignKey:
+            print(
+                level * ' ', f"CustomEndpoint alias: {self.alias} model: {self._model_class.__name__} primaryKey: {self.primaryKey} join_on: {self.foreignKey} parent: {parenName}")
+        else:
+            print(
+                level * ' ', f"CustomEndpoint alias: {self.alias} model: {self._model_class.__name__} primaryKey: {self.primaryKey} parent: {parenName}")
+        if isinstance(self.fields, tuple) and len(self.fields) > 0:
+            fields = self.getPrintableFields()
+            print(level * '  ', f"Fields: {fields}", sep=", ")
+        elif isinstance(self.fields, sqlalchemy.orm.attributes.InstrumentedAttribute):
+            print(level * '  ', f"Fields: {self.fields.key}", sep=", ")
+        if isinstance(self.children, CustomEndpoint):
+            self.children._parentResource = self
+            self.children._printIncludes(level + 1)
+        elif len(self.children) > 0:
+            for incl in self.children:
+                incl._parentResource = self
+                incl._printIncludes(level + 1)
+
+    def getPrintableFields(self):
+        result = ""
+        if len(self.fields) > 0:
+            for fld in self.fields:
+                if isinstance(fld, str):
+                    result += f" alias: {fld} "
+                elif isinstance(self.fields, sqlalchemy.orm.attributes.InstrumentedAttribute):
+                    result += f" {fld[0].key} " if isinstance(fld, tuple) else f" {fld} "
+        return result
+    
+    def _modifyRows(self, result):
+        """
+        Start at root row and descend to each child
+        Args:
+            result dict modified and shaped JSON
+        """
+        result[self.alias] = []
+        for row in self._dictRows:
+            newRow = self._modifyRow(row)
+            result[self.alias].append(newRow)
+            #need to link each newRow with one or more childRows
+            if isinstance(self.children, CustomEndpoint):
+                self.children._linkAndModifyRows(row, newRow)
+            elif len(self.children) > 0:
+                for child in self.children:
+                    child._linkAndModifyRows(row, newRow)
+
+    def _linkAndModifyRows(self, row: dict, modifiedRow: dict):
+        """
+            link rows to parent 
+        Args:
+            row (dict): this is the parent row
+            modifiedRow (dict): this is the same row that has been modified
+        """
+        if not self.isCombined:
+            modifiedRow[self.alias] = []
+        self._parentRow  = DotDict(row)
+        pkeyValue = row[self.foreignKey.key] if self.isParent and self.foreignKey.key in row else row[self.primaryKey]
+        fkey = self.primaryKey  if self.isParent and self.primaryKey in row else self.foreignKey.key if self.foreignKey is not None else None
+        for dictRow in self._dictRows:
+            if fkey is not None and f"{pkeyValue}" == f"{dictRow[fkey]}":
+                newRow = self._modifyRow(dictRow)
+                if self.isParent and self.isCombined:
+                    modifiedRow |= newRow
+                else:
+                    modifiedRow[self.alias].append(newRow)
+                if isinstance(self.children, CustomEndpoint):
+                    self.children._linkAndModifyRows(dictRow, newRow)
+                elif len(self.children) > 0:
+                    for include in self.children:
+                        include._linkAndModifyRows(dictRow, newRow)
+
+    def _modifyRow(self, dict_row: dict) -> dict:
+        #row = self.transform('LAC','',dict_row)
+        newRow = DotDict({})
+        tableRow = DotDict(dict_row)
+        if isinstance(self.fields, sqlalchemy.orm.attributes.InstrumentedAttribute):
+            f = self.fields
+            fieldName = f[0].key if isinstance(f, tuple) else f.key
+            alias = f[1] if isinstance(f, tuple) else fieldName
+            if fieldName in tableRow:
+                newRow[alias] = tableRow[fieldName]
+        elif len(self.fields) > 0:
+            for f in self.fields:
+                if isinstance(f,str):
+                    fieldName = f
+                    alias = fieldName
+                elif isinstance(f[0], sqlalchemy.sql. schema.Column):
+                    alias = f[0].description 
+                    fieldName = f[1] if isinstance(f, tuple) else fieldName
+                else:
+                    fieldName = f[0].key if isinstance(f, tuple) else f.key
+                    alias = f[1] if isinstance(f, tuple) else fieldName
+                if fieldName in tableRow:
+                    newRow[alias] = tableRow[fieldName]
+        else:
+            newRow = tableRow
+        # allow adding or changes using defined function
+        if self.calling is not None:
+            try:
+                resource_logger.debug(f"calling function {self.calling}")
+                self.calling(newRow, tableRow, self._parentRow)
+            except Exception as ex:
+                resource_logger.error(f"unable to execute fn {self.calling} error: {ex}")
+        if not self.isCombined:
+            self.insertCheckSum(newRow, tableRow)
+            
+        return newRow
+    
+    def insertCheckSum(self, newRow: dict, tableRow: dict):
+        if Args.opt_locking == "required" \
+            and ("S_CheckSum" not in newRow and "S_CheckSum" in tableRow):
+            newRow["S_CheckSum"] = tableRow.S_CheckSum
+            newRow = self.move_checksum(newRow)
+        elif "@metadata" in tableRow:
+            newRow["@metadata"] = tableRow["@metadata"]
+            if "S_CheckSum" in newRow:
+                newRow.pop("S_CheckSum")
+            
+    
+    def addRowToResult(self, result: any, rows: any):
+        """_summary_
+
+        Args:
+            result (any): this is the final output
+            rows (any): the unmodified dict rows
+        """
+        if self.foreignKey is None:
+            return
+        fkey = self.foreignKey.key if isinstance(self.foreignKey, object) else self.foreignKey
+        resource_logger.debug(f"Add Row to Result {self._model_class_name} using {fkey}")
+        keyList =  self._fkeyList if self.isParent else self._parentResource._pkeyList 
+        for parentKey in keyList:
+            for r in result:
+                r[self.alias] = []
+                for row in rows:
+                    fkeyValue = row.get(fkey)
+                    if parentKey == fkeyValue:
+                        #modifiedRow = self._modifyRow(row)
+                        r[self.alias].append(row)
+                        
+    def _populateResponse(self, jsonResponse):
+        """
+        Given a json response extract and populate internal dictRows
+        Args:
+            jsonResponse (_type_): _description_
+        """
+        #assume 1 data row
+        jsonDict = DotDict(jsonResponse)
+        if len(jsonDict.data) == 0:
+            return
+        for data in jsonDict.data:
+            key =  data["id"] 
+            row = data["attributes"]
+            if self.primaryKey == "id" and "id" not in row:
+                row["id"] = key #this is a hack since id is a jsonapi reserved value
+            self._dictRows.append(row) 
+            if key not in self._pkeyList:
+                self._pkeyList.append(key)
+            model_type = data["type"]
+            resource_logger.debug(f"_populateResponse row class on {self._model_class_name} using model_type: {model_type} with key {key}")
+        if self.children is not None:
+            included = jsonDict.included
+            if len(included) == 0:
+                return
+            if isinstance(self.children, list):
+                for child in self.children:
+                    child._parentResource = self
+                    child.processIncludedRows(included)
+            else:
+                self.children._parentResource = self
+                self.children.processIncludedRows(included)
+
+    def processIncludedRows(self, included: list):
+        for parentKey in self._parentResource._pkeyList:
+            for row in included:
+                model_class_name = row["type"]
+                if model_class_name == self._model_class_name:
+                    resource_logger.debug(f"includeRow for {self._model_class_name}")
+                    attrRow = row["attributes"]
+                    if "id" not in "attrs" and "id" in row:
+                        attrRow["id"] = row["id"]
+                    keyName = self.primaryKey if self.isParent else self.join_on.key
+                    if keyName in attrRow and parentKey == attrRow[keyName]:
+                        resource_logger.debug(f"includeRow for {self._model_class_name} checking {model_class_name} using Key: {keyName} ")
+                        #links = row["links"]
+                        #relns = row["relationships"]
+                        self._dictRows.append(attrRow)
+                        key = attrRow[self.primaryKey]
+                        if key not in self._pkeyList:
+                            self._pkeyList.append(key)
+        if self.children is not None:
+            if isinstance(self.children, list):
+                for child in self.children:
+                    child._parentResource = self
+                    child.processIncludedRows(included)
+            else:
+                self.children._parentResource = self
+                self.children.processIncludedRows(included)
+                
+    def insert_or_update(self, payload: dict,altKey: str = None) -> any: 
+        p = []
+        if not isinstance(payload, list):
+            p.append(payload)
+        else:
+            p = payload
+        result = []
+        for row in p:
+            clz = self.copy_dict_to_row(payload=row)
+            if altKey is not None:
+                setattr(clz, self.primaryKey, altKey)
+            session.add(clz)         
+            session.commit()
+            d = {}
+            for column in clz.__table__.columns:
+                d[column.name] = str(getattr(clz, column.name))
+            result.append(d)
+        return result
+    
+    def handlePayload(self, method: str, payload: any, url: str, jwt: str,altKey: str = None) -> any:
+        """ tests
+            stmt = ""
+            if method == 'POST':
+                stmt = insert(self._model_class).values(payload)
+            #elif  stmt = update(self._model_class).values(payload) #.where(self.primaryKey = 1)
+            clz = self._model_class
+            key = self.populateClass(clz, payload)
+            db.session.add(clz)
+            #db.engine.execute(stmt)
+            # db.session.select().filter_by().one()
+            return db.engine.execute(f"select * from {self._model_class_name} limit 1").one()
+        """
+        j = self.create_args(method, payload, altKey)
+        # check payload for a single row
+        clz = self._model_class
+        #key = self.populateClass(clz, payload)
+        print(self.to_dict(payload))
+        #  need to do a loop to post/patch each level
+        #pkey = None
+        #for row, model_clz in self.getRowFrom(payload, pkey): #cascade primary key
+        #    pkey = insert(self.model_clz).values(dmlColumnKeyMapping(row)).returning(model_clz.get(self.primary_key))
+        #    or
+        #    populate(model_clz, row, pkey)
+        #    session.add(model_clz)
+        #    pkey = model_clz.get(self.primary_key) if hassattr(model_clz, self.primary_key) else None
+        
+        key = altKey if altKey else payload[self.primaryKey] if self.primaryKey in payload else "-1"
+        if Args.security_enabled:
+            header = {"Authorization": jwt,"Content-Type": "application/json","accept": "application/vnd.api+json"}
+            response = (
+                requests.post(url=url, json=j, headers=header) #insert(self._model_class).values(self.dmlColumnKeyMapping(row)).returning("id")
+                if method == 'POST'
+                    else requests.patch(url=f"{url}/{key}", json=j, headers=header) #update(self._model_class).values(self.dmlColumnKeyMapping(row))
+            )
+        else:
+            response = {}
+            if method == "POST":
+                response = requests.post(url=url, json=j) 
+            elif method in ["PUT","PATCH" ]:
+                response = requests.patch(url=f"{url}/{key}", data=j) 
+        data  = json.loads(response.text)["data"]["attributes"]        
+        return json.dumps(data,indent=4, ensure_ascii=False).encode('utf8') if response.status_code < 301 else response.content
+
+    def populateClass(self, clz, payload):
+        for p in payload:
+            clz(p = payload[p])
+        return clz[self.primaryKey]
+    
+    def create_args(self, method:str, attributes:any, altKey:str = None):
+        key = altKey if altKey else attributes[self.primaryKey] if self.primaryKey in attributes else None
+        result = None
+        if key is None or method == 'POST':
+            result =  \
+                { "data": {
+                    "attributes": attributes,
+                    "type": self._model_class_name
+                }
+            }
+        else:
+            result = \
+                { "data": {
+                    "attributes": self.move_metadata(attributes),
+                    "type": self._model_class_name,
+                    "id": int(key) if self.primaryKeyType.python_type == int else key
+                }
+            }
+        v =  str(result)
+        v = v.replace("'","\"",1000)
+        return json.loads(v.replace("None","null",100))
+
+    def move_metadata(self, json_dict:dict) -> dict:
+        if "@metadata" in json_dict:
+            if json_dict["@metadata"]["checksum"] != 'override':
+                json_dict["S_CheckSum"] = json_dict["@metadata"]["checksum"]
+            json_dict.pop("@metadata")
+        return json_dict
+        
+    def quoteStr(self, val):
+        return val if f"{self.primaryKeyType}" == 'INTEGER' else f"'{val}'"
+    
+    def rows_to_dict(self: CustomEndpoint, result: flask_sqlalchemy.BaseQuery) -> list:
+        """
+        Converts SQLAlchemy result to dict array
+
+        Args:
+            result (object): SQLAlchemy result
+
+        Returns:
+            dict: dict array
+        """
+        from decimal import Decimal
+        import datetime 
+        rows = []
+        for each_row in result:
+            row_as_dict = {}
+            print(f'type(each_row): {type(each_row)}')
+            if isinstance (each_row, sqlalchemy.engine.row.Row):  # sqlalchemy.engine.row
+                row_as_dict = each_row._asdict()
+            else:
+                for a,v, in each_row.__dict__.items(): 
+                    if a != "_sa_instance_state":
+                        if isinstance(v, Decimal):
+                            row_as_dict[a] = str(v) 
+                        elif isinstance(v, datetime.date):
+                            row_as_dict[a] = v.strftime('%Y-%m-%d %H:%M:%S')
+                        else:
+                            row_as_dict[a] = v
+            if hasattr(each_row,"id"):
+                with contextlib.suppress(Exception):
+                    row_as_dict["id"] = each_row.id
+            rows.append(row_as_dict)
+        return rows
+
+    def row_to_dict(self: CustomEndpoint, row
+                , replace_attribute_tag: str = ""
+                , remove_links_relationships: bool = False) -> dict:
+        """
+        returns dict suitable for safrs response
+
+        Args:
+            row (safrs.DB.Model): a SQLAlchemy row
+            replace_attribute_tag (str): replace _attribute_ tag with this name
+            remove_links_relationships (bool): remove these tags
+        Returns:
+            _type_: dict (suitable for flask response)
+        """
+        row_as_dict = jsonify(row).json
+        resource_logger.debug(f'Row: {row_as_dict}')
+        if replace_attribute_tag != "":
+            row_as_dict[replace_attribute_tag] = row_as_dict.pop('attributes')
+        if remove_links_relationships:
+            row_as_dict.pop('links')
+            row_as_dict.pop('relationships')
+        if not hasattr(row_as_dict,"id"):
+            with contextlib.suppress(Exception):
+                row_as_dict["id"] = row["id"] 
+        return row_as_dict
+
+    def _processChildren(self):
+        resource_logger.debug(
+                f"_executeChildren a child: {self._model_class_name} isParent: {self.isParent}")
+        self._createRows()
+        self._executeChildren()
+        
+    def parseArgs(self,args):
+        '''
+        Args = filter.data.data.data = "someexpression=N" 
+        '''
+        tenant_filter = None
+        _filter = None
+        pkey = self.primaryKey
+        dots = self.getAlias()
+        filter_with_alias = f"filter.{dots}"if len(dots) > 0 else ""
+        value = args.get(pkey) if args.get(pkey) is not None else args.get(f"filter[{pkey}]")
+        #if value is None:
+        _sys_filter:str = args.get("sysfilter") 
+        _filter:str = args.get("filter") if args.get("filter") is not None else args.get(filter_with_alias) if args.get(filter_with_alias) is not None else None
+        
+        if _sys_filter:
+            if _sys_filter.startswith("equal("):
+                f = _sys_filter[6:-1].split(":")
+                pkey = f[0]
+                value = f[1]
+        elif _filter:
+            f = _filter.split("=")
+            if len(f) > 1 and f[1] != 'undefined' and f[0] == self.primaryKey:
+                q = '' if f[0] in ['OFFICEID','CUSTOMERID','ACCOUNTID','BRANCHID'] else "'" 
+                pkey = f'"{f[0]}"'
+                value = f"{q}{f[1]}{q}"
+
+        limit = args.get("page[limit]") or args.get("pagesize") or 20
+        offset = args.get("page[offset]") or args.get("offset")  or 0
+        sort = args.get("sort")
+        tenant_filter = _filter if  _filter is not None else "1=1"
+
+        return pkey, value, limit, offset, sort, tenant_filter
+    
+    def getAlias(self):
+        #fillter.data.data.data="someexpression=areaCode<3"
+        dots = ""
+        if self.children is not None:
+            if isinstance(self.children, CustomEndpoint):
+                return f"{self.alias}.{self.children.alias}"
+            for child in self.children:
+                if dots == "":
+                    dots = f".{self.alias}"
+                dots = f"{dots}.{child.alias}"
+        return dots
+        
+    def transform(self, style:str, key:str, json_: dict) -> dict:
+	# use this to change the output (pipeline) of the result
+        json_dict = {}
+        json_result = []
+        result = []
+        try:
+            if self._method == 'OPTIONS':
+                return json_
+            #TODO - fixup asscii to utf-8
+            json_dict = json.loads(json_) if isinstance(json_, bytes) else json_
+            json_result = json_dict.get(key, json_dict) if key in json_dict else json_dict if isinstance(json_dict, list) else [json_dict]
+        except Exception as ex:
+            resource_logger.error(f"Transform Error on style {style} using key: {key} on {json_} error: {ex}")
+            return json_
+
+        
+        if isinstance(json_result,list):
+            newRes = []
+            for row in json_result:
+                r = self.move_checksum(row)
+                newRes.append(r)
+            result = newRes
+        result = self.move_checksum(json_result)
+        result if isinstance(result,list) else [result]
+        
+        if style == "JSONAPI":
+            # Ontimize using the JSONAPI with the API Bridge
+            data = []
+            for row in result:
+                pkey = row[self.primaryKey] if self.primaryKey in row else None
+                if pkey is None:
+                    pkey = row["id"] if "id" in row else row["Id"] if "Id" in row else None
+                data.append({"attributes": row,"type": self._model_class_name, "id": pkey })
+            data = {"data": data,
+                "meta": {
+                    "count": len(result),
+                    "limit": self.pagesize,
+                    "total": len(result)
+                }        
+            }
+            result = data
+        elif style == "OntimizeEE":
+            #API Bridge - lets Ontimize work out-of-the-box
+            recordsNumber = self.totalQueryRecordsNumber #if len(result) == 0 else self.startRecordIndex
+            startRecord = self.startRecordIndex
+            result = {"code":0,"totalQueryRecordsNumber": recordsNumber, "startRecordIndex": startRecord, "message":"ApiLogicServer","data": result ,"sqlTypes":{}}
+        #if style == "LAC": default
+        return result
+    
+
+    def move_checksum(self, json_dict:any) -> dict:
+        if isinstance(json_dict, dict):
+            if "S_CheckSum" in json_dict:
+                checksum = json_dict["S_CheckSum"]
+                pk = json_dict[self.primaryKey] if self.primaryKey in json_dict else ""
+                href = f"{self._href}/{pk}" if self._href is not None else ""
+                json_dict["@metadata"] = { "checksum" : checksum, "href": href}
+                json_dict.pop("S_CheckSum")
+            if "_check_sum_" in json_dict:
+                json_dict.pop("_check_sum_")
+        elif isinstance(json_dict,list):
+            for json_ in json_dict:
+                self.move_checksum(json_)
+        if self.children is not None:
+            if isinstance(self.children, list):
+                for child in self.children:
+                    if child.alias in json_dict:
+                        child.move_checksum(json_dict[child.alias])
+            else:
+                if self.children.alias in json_dict:
+                        self.children.move_checksum(json_dict[self.children.alias])
+        return json_dict
+    
+    def to_dict(self, row: object, current_endpoint: 'CustomEndpoint' = None) -> dict:
+        """returns row as dict per custom resource definition, with subobjects
+
+        Args:
+            row (_type_): a SQLAlchemy row
+
+        Returns:
+            dict: row formatted as dict
+        """
+        custom_endpoint = self
+        if current_endpoint is not None:
+            custom_endpoint = current_endpoint
+        # row_as_dict = self.row_to_dict(row)
+        row_as_dict = {}
+        for each_field in custom_endpoint.fields:
+            if isinstance(each_field, tuple):
+                row_as_dict[each_field[1]] = getattr(row, each_field[0].name)
+            else:
+                if isinstance(each_field, str):
+                    print("Coding error - you need to use TUPLE for attr/alias")
+                row_as_dict[each_field.name] = getattr(row, each_field.name)
+        
+        custom_endpoint_child_list = custom_endpoint.children
+        if isinstance(custom_endpoint_child_list, list) is False:
+            custom_endpoint_child_list = []
+            custom_endpoint_child_list.append(custom_endpoint.children)
+        for each_child_def in custom_endpoint_child_list:
+            child_property_name = each_child_def.role_name
+            if child_property_name == '':
+                child_property_name = "OrderList"  # TODO default from class name
+            if child_property_name.startswith('Product'):
+                debug = 'good breakpoint'
+            row_dict_child_list = getattr(row, child_property_name)
+            row_as_dict[each_child_def.alias] = []
+            if each_child_def.isParent:
+                the_parent = getattr(row, child_property_name)
+                the_parent_to_dict = self.to_dict(row = the_parent, current_endpoint = each_child_def)
+                row_as_dict[each_child_def.alias].append(the_parent_to_dict)
+            else:
+                for each_child in row_dict_child_list:
+                    each_child_to_dict = self.to_dict(row = each_child, current_endpoint = each_child_def)
+                    row_as_dict[each_child_def.alias].append(each_child_to_dict)
+        return row_as_dict
+    
+
+    def to_row(self, row_dict: dict, current_endpoint: 'CustomEndpoint' = None) -> object:
+        """Returns SQLAlchemy row(s), converted from safrs-request per subclass custom resource
+
+        Args:
+            safrs_request: a 
+            current_endpoint (IntegrationEndpoint, optional): _description_. Defaults to None.
+
+        Returns:
+            object: SQLAlchemy row / sub-rows, ready to insert
+        """
+
+        print( f"to_row receives row_dict: {row_dict}" )
+
+        custom_endpoint = self
+        if current_endpoint is not None:
+            custom_endpoint = current_endpoint
+        sql_alchemy_row = custom_endpoint._model_class()     # new instance
+        for each_field in custom_endpoint.fields:           # attr mapping  TODO 1 field, not array
+            if isinstance(each_field, tuple):
+                setattr(sql_alchemy_row, each_field[0].name, row_dict[each_field[1]])
+            else:
+                if isinstance(each_field, str):
+                    print("Coding error - you need to use TUPLE for attr/alias")
+                setattr(sql_alchemy_row, each_field.name, row_dict[each_field.name])
+        row_dict = self.move_metadata(row_dict) ## Validate TODO
+        custom_endpoint_child_list = custom_endpoint.children
+        if isinstance(custom_endpoint_child_list, list) is False:
+            custom_endpoint_child_list = []
+            custom_endpoint_child_list.append(custom_endpoint.children)
+        for each_child_def in custom_endpoint_child_list:
+            child_property_name = each_child_def.alias
+            if child_property_name.startswith('Items'):
+                debug = 'good breakpoint'
+            if child_property_name in row_dict:
+                row_dict_child_list = row_dict[child_property_name]
+                # row_as_dict[each_child_def.alias] = []  # set up row_dict child array
+                if each_child_def.isParent:
+                    #the_parent = getattr(row, child_property_name)
+                    #the_parent_to_dict = self.to_dict(row = the_parent, current_endpoint = each_child_def)
+                    #row_as_dict[each_child_def.alias].append(the_parent_to_dict)
+                    pass
+                else:
+                    for each_row_dict_child in row_dict_child_list:  # recurse for each_child
+                        each_child_row = self.to_row(row_dict = each_row_dict_child, current_endpoint = each_child_def)
+                        child_list = getattr(sql_alchemy_row, each_child_def.role_name)
+                        child_list.append(each_child_row)
+        return sql_alchemy_row
+    
+    def modifyPath(self, path):
+        p = path.split("/")
+        return path.replace(p[len(p)-1],"")
+    
+    def insert(self, request: any, payload: dict) -> any:
+        sqlalchemy_row = self.copy_dict_to_row(payload)
+        session.add(sqlalchemy_row)
+        session.commit()
+        session.flush()
+        return sqlalchemy_row
+    
+    def update(self, payload: dict, pkey: any) -> any:
+        payload = self.move_metadata(payload)
+        sqlalchemy_row = self.copy_dict_to_row(payload)
+        session.add(sqlalchemy_row)
+        session.commit()
+        session.flush()
+        return sqlalchemy_row
+    
+    def copy_dict_to_row(self, payload: dict) -> any:
+        sql_alchemy_row = self._model_class()   
+        if "@metadata" in payload:
+            payload.pop("@metadata")
+        for v in payload:
+            if v in self._columnNames:
+                setattr(sql_alchemy_row, v , payload[v])
+        return sql_alchemy_row
+        
+    def transform_to_safrs(self, payload: any, pkey: any = None):
+        if pkey is None:
+            return {"data":
+                {
+                    "attributes": payload
+                },
+                "type" : self._model_class_name
+            }
+        else:
+            return {"data":
+                {
+                    "attributes": payload
+                },
+                "type" : self._model_class_name,
+                "id": pkey
+            }