* Make extension points abstract classes.

* Improve ML.NET version handling to actually rely on the assembly vs. a constant string we have to change.
* Tighten checks of arguments to the context implementation.

src/Microsoft.ML.Data/Model/Onnx/ICanSaveOnnx.cs

@@ -27,7 +27,7 @@ public interface ITransformCanSaveOnnx: ICanSaveOnnx, IDataTransform
         /// Save as ONNX.
         /// </summary>
         /// <param name="ctx">The ONNX program being built</param>
-        void SaveAsOnnx(IOnnxContext ctx);
+        void SaveAsOnnx(OnnxContext ctx);
     }
 
     /// <summary>
@@ -52,7 +52,7 @@ public interface IBindableCanSaveOnnx : ICanSaveOnnx, ISchemaBindableMapper
         /// the outputs produced by this bindable mapper. This is the array that holds
         /// those names, so that implementors of this method know what to produce in
         /// <paramref name="ctx"/>.</param>
-        bool SaveAsOnnx(IOnnxContext ctx, RoleMappedSchema schema, string[] outputNames);
+        bool SaveAsOnnx(OnnxContext ctx, RoleMappedSchema schema, string[] outputNames);
     }
 
     /// <summary>
@@ -61,7 +61,7 @@ public interface IBindableCanSaveOnnx : ICanSaveOnnx, ISchemaBindableMapper
     /// </summary>
     public interface ISingleCanSaveOnnx : ICanSaveOnnx
     {
-        bool SaveAsOnnx(IOnnxContext ctx, string[] outputNames, string featureColumn);
+        bool SaveAsOnnx(OnnxContext ctx, string[] outputNames, string featureColumn);
     }
 
     /// <summary>
@@ -70,6 +70,6 @@ public interface ISingleCanSaveOnnx : ICanSaveOnnx
     /// </summary>
     public interface IDistCanSaveOnnx : ISingleCanSaveOnnx, IValueMapperDist
     {
-        new bool SaveAsOnnx(IOnnxContext ctx, string[] outputNames, string featureColumn);
+        new bool SaveAsOnnx(OnnxContext ctx, string[] outputNames, string featureColumn);
     }
 }

src/Microsoft.ML.Data/Model/Onnx/IOnnxContext.cs → src/Microsoft.ML.Data/Model/Onnx/OnnxContext.cs

@@ -16,22 +16,22 @@ namespace Microsoft.ML.Runtime.Model.Onnx
     /// 
     /// 
     /// </summary>
-    public interface IOnnxContext
+    public abstract class OnnxContext
     {
         /// <summary>
         /// Generates a unique name for the node based on a prefix.
         /// </summary>
         /// <param name="prefix">The prefix for the node</param>
         /// <returns>A name that has not yet been returned from this function, starting with <paramref name="prefix"/></returns>
-        string GetNodeName(string prefix);
+        public abstract string GetNodeName(string prefix);
 
         /// <summary>
         /// Looks up whether a given data view column has a mapping in the ONNX context. Once confirmed, callers can
         /// safely call <see cref="GetVariableName(string)"/>.
         /// </summary>
         /// <param name="colName">The data view column name</param>
         /// <returns>Whether the column is mapped in this context</returns>
-        bool ContainsColumn(string colName);
+        public abstract bool ContainsColumn(string colName);
 
         /// <summary>
         /// Stops tracking a column.
@@ -40,7 +40,7 @@ public interface IOnnxContext
         /// <param name="removeVariable">Remove associated ONNX variable. This is useful in the event where an output
         /// variable is created through <see cref="AddIntermediateVariable(ColumnType, string, bool)"/>before realizing
         /// the transform cannot actually save as ONNX.</param>
-        void RemoveColumn(string colName, bool removeVariable = false);
+        public abstract void RemoveColumn(string colName, bool removeVariable = false);
 
         /// <summary>
         /// Removes an ONNX variable. If removeColumn is true then it also removes the tracking for the <see
@@ -49,7 +49,7 @@ public interface IOnnxContext
         /// <param name="variableName">ONNX variable to remove. Note that this is an ONNX variable name, not an <see
         /// cref="IDataView"/> column name</param>
         /// <param name="removeColumn">IDataView column to stop tracking</param>
-        void RemoveVariable(string variableName, bool removeColumn);
+        public abstract void RemoveVariable(string variableName, bool removeColumn);
 
         /// <summary>
         /// ONNX variables are referred to by name. At each stage of a ML.NET pipeline, the corresponding
@@ -61,7 +61,7 @@ public interface IOnnxContext
         /// </summary>
         /// <param name="colName">The data view column name</param>
         /// <returns>The ONNX variable name corresponding to that data view column</returns>
-        string GetVariableName(string colName);
+        public abstract string GetVariableName(string colName);
 
         /// <summary>
         /// Establishes a new mapping from an data view column in the context, if necessary generates a unique name, and
@@ -72,7 +72,7 @@ public interface IOnnxContext
         /// <param name="skip">Whether we should skip the process of establishing the mapping from data view column to
         /// ONNX variable name.</param>
         /// <returns>The returned value is the name of the variable corresponding </returns>
-        string AddIntermediateVariable(ColumnType type, string colName, bool skip = false);
+        public abstract string AddIntermediateVariable(ColumnType type, string colName, bool skip = false);
 
         /// <summary>
         /// Creates an ONNX node 
@@ -84,25 +84,25 @@ public interface IOnnxContext
         /// <param name="name">The name of the operator, which ought to be something returned from <see cref="GetNodeName(string)"/></param>
         /// <param name="domain">The domain of the ONNX operator, if non-default</param>
         /// <returns>A node added to the in-progress ONNX graph, that attributes can be set on</returns>
-        IOnnxNode CreateNode(string opType, IEnumerable<string> inputs,
+        public abstract OnnxNode CreateNode(string opType, IEnumerable<string> inputs,
             IEnumerable<string> outputs, string name, string domain = null);
     }
 
     public static class OnnxContextExtensions
     {
         /// <summary>
-        /// Convenience alternative to <see cref="IOnnxContext.CreateNode(string, IEnumerable{string}, IEnumerable{string}, string, string)"/>
+        /// Convenience alternative to <see cref="OnnxContext.CreateNode(string, IEnumerable{string}, IEnumerable{string}, string, string)"/>
         /// for the case where there is exactly one input and output.
         /// </summary>
         /// <param name="ctx">The ONNX save context</param>
         /// <param name="opType">The name of the ONNX operator to apply</param>
         /// <param name="input">The name of the variable as input</param>
         /// <param name="output">The name of the variable as output,
-        /// which ought to have been something returned from <see cref="IOnnxContext.AddIntermediateVariable(ColumnType, string, bool)"/></param>
-        /// <param name="name">The name of the operator, which ought to be something returned from <see cref="IOnnxContext.GetNodeName(string)"/></param>
+        /// which ought to have been something returned from <see cref="OnnxContext.AddIntermediateVariable(ColumnType, string, bool)"/></param>
+        /// <param name="name">The name of the operator, which ought to be something returned from <see cref="OnnxContext.GetNodeName(string)"/></param>
         /// <param name="domain">The domain of the ONNX operator, if non-default</param>
         /// <returns>A node added to the in-progress ONNX graph, that attributes can be set on</returns>
-        public static IOnnxNode CreateNode(this IOnnxContext ctx,
+        public static OnnxNode CreateNode(this OnnxContext ctx,
             string opType, string input, string output, string name, string domain = null)
             => ctx.CreateNode(opType, new[] { input }, new[] { output }, name, domain);
     }

src/Microsoft.ML.Data/Model/Onnx/OnnxNode.cs

@@ -0,0 +1,32 @@
+// Licensed to the .NET Foundation under one or more agreements.
+// The .NET Foundation licenses this file to you under the MIT license.
+// See the LICENSE file in the project root for more information.
+
+using System.Collections.Generic;
+using Microsoft.ML.Runtime.Data;
+
+namespace Microsoft.ML.Runtime.Model.Onnx
+{
+    /// <summary>
+    /// An abstraction for an ONNX node as created by
+    /// <see cref="OnnxContext.CreateNode(string, IEnumerable{string}, IEnumerable{string}, string, string)"/>.
+    /// That method creates a with inputs and outputs, but this object can modify the node further
+    /// by adding attributes (in ONNX parlance, attributes are more or less constant parameterizations).
+    /// </summary>
+    public abstract class OnnxNode
+    {
+        public abstract void AddAttribute(string argName, double value);
+        public abstract void AddAttribute(string argName, long value);
+        public abstract void AddAttribute(string argName, DvText value);
+        public abstract void AddAttribute(string argName, string value);
+        public abstract void AddAttribute(string argName, bool value);
+
+        public abstract void AddAttribute(string argName, IEnumerable<double> value);
+        public abstract void AddAttribute(string argName, IEnumerable<float> value);
+        public abstract void AddAttribute(string argName, IEnumerable<long> value);
+        public abstract void AddAttribute(string argName, IEnumerable<DvText> value);
+        public abstract void AddAttribute(string argName, string[] value);
+        public abstract void AddAttribute(string argName, IEnumerable<string> value);
+        public abstract void AddAttribute(string argName, IEnumerable<bool> value);
+    }
+}

src/Microsoft.ML.Data/Prediction/Calibrator.cs

@@ -296,7 +296,7 @@ public void SaveAsPfa(BoundPfaContext ctx, JToken input,
             probToken = ctx.DeclareVar(prob, probExpression);
         }
 
-        public bool SaveAsOnnx(IOnnxContext ctx, string[] outputNames, string featureColumnName)
+        public bool SaveAsOnnx(OnnxContext ctx, string[] outputNames, string featureColumnName)
         {
             Host.CheckValue(ctx, nameof(ctx));
             Host.CheckValue(outputNames, nameof(outputNames));
@@ -658,7 +658,7 @@ public void SaveAsPfa(BoundPfaContext ctx, RoleMappedSchema schema, string[] out
             ctx.Hide(outputs);
         }
 
-        public bool SaveAsOnnx(IOnnxContext ctx, RoleMappedSchema schema, string[] outputs)
+        public bool SaveAsOnnx(OnnxContext ctx, RoleMappedSchema schema, string[] outputs)
         {
             Host.CheckValue(ctx, nameof(ctx));
             Host.CheckParam(Utils.Size(outputs) == 2, nameof(outputs), "Expected this to have two outputs");
@@ -1429,7 +1429,7 @@ public JToken SaveAsPfa(BoundPfaContext ctx, JToken input)
                 PfaUtils.Call("+", -ParamB, PfaUtils.Call("*", -ParamA, input)));
         }
 
-        public bool SaveAsOnnx(IOnnxContext ctx, string[] scoreProbablityColumnNames, string featureColumnName)
+        public bool SaveAsOnnx(OnnxContext ctx, string[] scoreProbablityColumnNames, string featureColumnName)
         {
             _host.CheckValue(ctx, nameof(ctx));
             _host.CheckValue(scoreProbablityColumnNames, nameof(scoreProbablityColumnNames));

src/Microsoft.ML.Data/Scorers/BinaryClassifierScorer.cs

@@ -184,7 +184,7 @@ protected override void SaveCore(ModelSaveContext ctx)
             ctx.Writer.Write(_threshold);
         }
 
-        public override void SaveAsOnnx(IOnnxContext ctx)
+        public override void SaveAsOnnx(OnnxContext ctx)
         {
             Host.CheckValue(ctx, nameof(ctx));
             Host.Assert(Bindable is IBindableCanSaveOnnx);

src/Microsoft.ML.Data/Scorers/GenericScorer.cs

@@ -214,7 +214,7 @@ public void SaveAsPfa(BoundPfaContext ctx)
             pfaBindable.SaveAsPfa(ctx, schema, outColNames);
         }
 
-        public void SaveAsOnnx(IOnnxContext ctx)
+        public void SaveAsOnnx(OnnxContext ctx)
         {
             Host.CheckValue(ctx, nameof(ctx));
             Host.Assert(Bindable is IBindableCanSaveOnnx);

src/Microsoft.ML.Data/Scorers/MultiClassClassifierScorer.cs

@@ -203,7 +203,7 @@ public void SaveAsPfa(BoundPfaContext ctx, RoleMappedSchema schema, string[] out
                 ((IBindableCanSavePfa)_bindable).SaveAsPfa(ctx, schema, outputNames);
             }
 
-            public bool SaveAsOnnx(IOnnxContext ctx, RoleMappedSchema schema, string[] outputNames)
+            public bool SaveAsOnnx(OnnxContext ctx, RoleMappedSchema schema, string[] outputNames)
             {
                 Contracts.CheckValue(ctx, nameof(ctx));
                 Contracts.CheckValue(schema, nameof(schema));

src/Microsoft.ML.Data/Scorers/PredictedLabelScorerBase.cs

@@ -365,7 +365,7 @@ public void SaveAsPfa(BoundPfaContext ctx)
 
         protected abstract JToken PredictedLabelPfa(string[] mapperOutputs);
 
-        public virtual void SaveAsOnnx(IOnnxContext ctx)
+        public virtual void SaveAsOnnx(OnnxContext ctx)
         {
             Host.CheckValue(ctx, nameof(ctx));
             Host.Assert(Bindable is IBindableCanSaveOnnx);

src/Microsoft.ML.Data/Scorers/SchemaBindablePredictorWrapper.cs

@@ -99,7 +99,7 @@ public virtual void SaveAsPfa(BoundPfaContext ctx, RoleMappedSchema schema, stri
             ctx.Hide(outputNames);
         }
 
-        public virtual bool SaveAsOnnx(IOnnxContext ctx, RoleMappedSchema schema, string[] outputNames) => false;
+        public virtual bool SaveAsOnnx(OnnxContext ctx, RoleMappedSchema schema, string[] outputNames) => false;
 
         public ISchemaBoundMapper Bind(IHostEnvironment env, RoleMappedSchema schema)
         {
@@ -289,7 +289,7 @@ public override void SaveAsPfa(BoundPfaContext ctx, RoleMappedSchema schema, str
             ctx.DeclareVar(outputNames[0], scoreToken);
         }
 
-        public override bool SaveAsOnnx(IOnnxContext ctx, RoleMappedSchema schema, string[] outputNames)
+        public override bool SaveAsOnnx(OnnxContext ctx, RoleMappedSchema schema, string[] outputNames)
         {
             Contracts.CheckValue(ctx, nameof(ctx));
             Contracts.CheckValue(schema, nameof(schema));
@@ -403,7 +403,7 @@ public override void SaveAsPfa(BoundPfaContext ctx, RoleMappedSchema schema, str
             Contracts.Assert(ctx.TokenOrNullForName(outputNames[1]) == probToken.ToString());
         }
 
-        public override bool SaveAsOnnx(IOnnxContext ctx, RoleMappedSchema schema, string[] outputNames)
+        public override bool SaveAsOnnx(OnnxContext ctx, RoleMappedSchema schema, string[] outputNames)
         {
             Contracts.CheckValue(ctx, nameof(ctx));
             Contracts.CheckValue(schema, nameof(schema));

src/Microsoft.ML.Data/Transforms/ConcatTransform.cs

@@ -687,7 +687,7 @@ public void SaveAsPfa(BoundPfaContext ctx)
             ctx.DeclareVar(toDeclare.ToArray());
         }
 
-        public void SaveAsOnnx(IOnnxContext ctx)
+        public void SaveAsOnnx(OnnxContext ctx)
         {
             Host.CheckValue(ctx, nameof(ctx));
             Host.Assert(CanSaveOnnx);

src/Microsoft.ML.Data/Transforms/KeyToVectorTransform.cs

@@ -241,7 +241,7 @@ protected override JToken SaveAsPfaCore(BoundPfaContext ctx, int iinfo, ColInfo
                 PfaUtils.Call("cast.fanoutDouble", -1, 0, keyCount, false), PfaUtils.FuncRef("u." + funcName));
         }
 
-        protected override bool SaveAsOnnxCore(IOnnxContext ctx, int iinfo, ColInfo info, string srcVariableName, string dstVariableName)
+        protected override bool SaveAsOnnxCore(OnnxContext ctx, int iinfo, ColInfo info, string srcVariableName, string dstVariableName)
         {
             string opType = "OneHotEncoder";
             var node = ctx.CreateNode(opType, srcVariableName, dstVariableName, ctx.GetNodeName(opType));

src/Microsoft.ML.Data/Transforms/NormalizeColumn.cs

@@ -435,7 +435,7 @@ private AffineColumnFunction(IHost host)
 
             public abstract JToken PfaInfo(BoundPfaContext ctx, JToken srcToken);
             public bool CanSaveOnnx => true;
-            public abstract bool OnnxInfo(IOnnxContext ctx, IOnnxNode nodeProtoWrapper, int featureCount);
+            public abstract bool OnnxInfo(OnnxContext ctx, OnnxNode nodeProtoWrapper, int featureCount);
 
             public abstract Delegate GetGetter(IRow input, int icol);
 
@@ -548,7 +548,7 @@ public JToken PfaInfo(BoundPfaContext ctx, JToken srcToken)
 
             public bool CanSaveOnnx => false;
 
-            public bool OnnxInfo(IOnnxContext ctx, IOnnxNode nodeProtoWrapper, int featureCount)
+            public bool OnnxInfo(OnnxContext ctx, OnnxNode nodeProtoWrapper, int featureCount)
                 => throw Host.ExceptNotSupp();
 
             public abstract Delegate GetGetter(IRow input, int icol);
@@ -673,7 +673,7 @@ public JToken PfaInfo(BoundPfaContext ctx, JToken srcToken)
 
             public bool CanSaveOnnx => false;
 
-            public bool OnnxInfo(IOnnxContext ctx, IOnnxNode nodeProtoWrapper, int featureCount)
+            public bool OnnxInfo(OnnxContext ctx, OnnxNode nodeProtoWrapper, int featureCount)
                 => throw Host.ExceptNotSupp();
 
             public abstract Delegate GetGetter(IRow input, int icol);

src/Microsoft.ML.Data/Transforms/NormalizeColumnDbl.cs

@@ -577,7 +577,7 @@ public override void Save(ModelSaveContext ctx)
                     public override JToken PfaInfo(BoundPfaContext ctx, JToken srcToken)
                         => PfaUtils.Call("*", PfaUtils.Call("-", srcToken, Offset), Scale);
 
-                    public override bool OnnxInfo(IOnnxContext ctx, IOnnxNode nodeProtoWrapper, int featureCount)
+                    public override bool OnnxInfo(OnnxContext ctx, OnnxNode nodeProtoWrapper, int featureCount)
                     {
                         nodeProtoWrapper.AddAttribute("offset", Enumerable.Repeat(Offset, featureCount));
                         nodeProtoWrapper.AddAttribute("scale", Enumerable.Repeat(Scale, featureCount));
@@ -648,7 +648,7 @@ public override JToken PfaInfo(BoundPfaContext ctx, JToken srcToken)
                         return PfaUtils.Call("a.zipmap", srcToken, scaleCell, PfaUtils.FuncRef(ctx.Pfa.EnsureMul(itemType)));
                     }
 
-                    public override bool OnnxInfo(IOnnxContext ctx, IOnnxNode node, int featureCount)
+                    public override bool OnnxInfo(OnnxContext ctx, OnnxNode node, int featureCount)
                     {
 
                         if (Offset != null)