Require constant() for large constants and fix a type resolving bug (#9769)

* Require constant() for large constants and fix a type resolving bug

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

* Remove noise

1 files changed, 93 insertions, 0 deletions

diff --git a/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/TensorMaxMinTransformer.java b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/TensorMaxMinTransformer.java
new file mode 100644
index 00000000000..979c5b0f88c
--- /dev/null
+++ b/searchlib/src/main/java/com/yahoo/searchlib/rankingexpression/transform/TensorMaxMinTransformer.java
@@ -0,0 +1,93 @@
+// Copyright 2017 Yahoo Holdings. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
+package com.yahoo.searchlib.rankingexpression.transform;
+
+import com.yahoo.searchlib.rankingexpression.Reference;
+import com.yahoo.searchlib.rankingexpression.rule.CompositeNode;
+import com.yahoo.searchlib.rankingexpression.rule.ExpressionNode;
+import com.yahoo.searchlib.rankingexpression.rule.FunctionNode;
+import com.yahoo.searchlib.rankingexpression.rule.NameNode;
+import com.yahoo.searchlib.rankingexpression.rule.ReferenceNode;
+import com.yahoo.searchlib.rankingexpression.rule.TensorFunctionNode;
+import com.yahoo.tensor.TensorType;
+import com.yahoo.tensor.evaluation.TypeContext;
+import com.yahoo.tensor.functions.Reduce;
+
+import java.util.Optional;
+
+/**
+ * Transforms min(tensor,dim) and max(tensor,dim) to
+ * reduce(tensor,min/max,dim). This is necessary as the backend does
+ * not recognize these forms of min and max.
+ *
+ * @author lesters
+ */
+public class TensorMaxMinTransformer<CONTEXT extends TransformContext> extends ExpressionTransformer<CONTEXT> {
+
+    @Override
+    public ExpressionNode transform(ExpressionNode node, CONTEXT context) {
+        if (node instanceof CompositeNode) {
+            node = transformChildren((CompositeNode) node, context);
+        }
+        if (node instanceof FunctionNode) {
+            node = transformFunctionNode((FunctionNode) node, context.types());
+        }
+        return node;
+    }
+
+    public static ExpressionNode transformFunctionNode(FunctionNode node, TypeContext<Reference> context) {
+        switch (node.getFunction()) {
+            case min:
+            case max:
+                return transformMaxAndMinFunctionNode(node, context);
+        }
+        return node;
+    }
+
+    /**
+     * Transforms max and min functions if the first
+     * argument returns a tensor type and the second argument is a valid
+     * dimension in the tensor.
+     */
+    private static ExpressionNode transformMaxAndMinFunctionNode(FunctionNode node, TypeContext<Reference> context) {
+        if (node.children().size() != 2) {
+            return node;
+        }
+        ExpressionNode arg1 = node.children().get(0);
+        Optional<String> dimension = dimensionName(node.children().get(1));
+        if (dimension.isPresent()) {
+            TensorType type = arg1.type(context);
+            if (type.dimension(dimension.get()).isPresent()) {
+                return replaceMaxAndMinFunction(node);
+            }
+        }
+        return node;
+    }
+
+    private static Optional<String> dimensionName(ExpressionNode node) {
+        if (node instanceof ReferenceNode) {
+            Reference reference = ((ReferenceNode)node).reference();
+            if (reference.isIdentifier())
+                return Optional.of(reference.name());
+            else
+                return Optional.empty();
+        }
+        else if (node instanceof NameNode) {
+            return Optional.of(((NameNode)node).getValue());
+        }
+        else {
+            return Optional.empty();
+        }
+    }
+
+    private static ExpressionNode replaceMaxAndMinFunction(FunctionNode node) {
+        ExpressionNode arg1 = node.children().get(0);
+        ExpressionNode arg2 = node.children().get(1);
+
+        TensorFunctionNode.TensorFunctionExpressionNode expression = TensorFunctionNode.wrapArgument(arg1);
+        Reduce.Aggregator aggregator = Reduce.Aggregator.valueOf(node.getFunction().name());
+        String dimension = ((ReferenceNode) arg2).getName();
+
+        return new TensorFunctionNode(new Reduce(expression, aggregator, dimension));
+    }
+
+}