Allow the ndarray type to carry a shape.

include/npcomp/Dialect/Numpy/IR/NumpyDialect.h

@@ -49,10 +49,21 @@ class NdArrayType
 public:
   using Base::Base;
   static bool kindof(unsigned kind) { return kind == NumpyTypes::NdArray; }
-  static NdArrayType get(Type optionalDtype);
+
+  /// Constructs an NdArray with a dtype and no shape. Setting the dtype
+  /// to !basicpy.UnknownType will print as ?.
+  static NdArrayType get(Type dtype,
+                         llvm::Optional<ArrayRef<int64_t>> shape = llvm::None);
+
+  /// Returns whether the dtype is a concrete type (versus
+  /// !basicpy.UnknownType).
   bool hasKnownDtype();
   Type getDtype();
 
+  /// If the shape has been partially specified, this will have a value.
+  /// unknown dimensions are -1.
+  llvm::Optional<ArrayRef<int64_t>> getOptionalShape();
+
   // CPA::TypeMapInterface methods.
   Typing::CPA::TypeNode *mapToCPAType(Typing::CPA::Context &context);
 };

lib/Dialect/Numpy/IR/NumpyDialect.cpp

@@ -35,19 +35,71 @@ Type NumpyDialect::parseType(DialectAsmParser &parser) const {
     return AnyDtypeType::get(getContext());
   if (keyword == "ndarray") {
     // Parse:
-    //   ndarray<?>
-    //   ndarray<i32>
+    //   ndarray<*:?>
+    //   ndarray<*:i32>
+    //   ndarary<[1,2,3]:i32>
+    // Note that this is a different syntax than the built-ins as the dialect
+    // parser is not general enough to parse a dimension list with an optional
+    // element type (?). The built-in form is also remarkably ambiguous when
+    // considering extending it.
     Type dtype = Basicpy::UnknownType::get(getContext());
+    bool hasShape = false;
+    llvm::SmallVector<int64_t, 4> shape;
     if (parser.parseLess())
       return Type();
-    if (failed(parser.parseOptionalQuestion())) {
-      // Specified dtype.
-      if (parser.parseType(dtype))
+    if (succeeded(parser.parseOptionalStar())) {
+      // Unranked.
+    } else {
+      // Parse dimension list.
+      hasShape = true;
+      if (parser.parseLSquare())
+        return Type();
+      for (bool first = true;; first = false) {
+        if (!first) {
+          if (failed(parser.parseOptionalComma())) {
+            break;
+          }
+        }
+        if (succeeded(parser.parseOptionalQuestion())) {
+          shape.push_back(-1);
+          continue;
+        }
+
+        int64_t dim;
+        auto optionalPr = parser.parseOptionalInteger(dim);
+        if (optionalPr.hasValue()) {
+          if (failed(*optionalPr))
+            return Type();
+          shape.push_back(dim);
+          continue;
+        }
+        break;
+      }
+      if (parser.parseRSquare()) {
         return Type();
       }
-    if (parser.parseGreater())
+    }
+
+    // Parse colon dtype.
+    if (parser.parseColon()) {
       return Type();
-    return NdArrayType::get(dtype);
+    }
+
+    if (failed(parser.parseOptionalQuestion())) {
+      // Specified dtype.
+      if (parser.parseType(dtype)) {
+        return Type();
+      }
+    }
+    if (parser.parseGreater()) {
+      return Type();
+    }
+
+    llvm::Optional<ArrayRef<int64_t>> optionalShape;
+    if (hasShape)
+      optionalShape = shape;
+    auto ndarray = NdArrayType::get(dtype, optionalShape);
+    return ndarray;
   }
 
   parser.emitError(parser.getNameLoc(), "unknown numpy type: ") << keyword;
@@ -62,8 +114,23 @@ void NumpyDialect::printType(Type type, DialectAsmPrinter &os) const {
   case NumpyTypes::NdArray: {
     auto unknownType = Basicpy::UnknownType::get(getContext());
     auto ndarray = type.cast<NdArrayType>();
+    auto shape = ndarray.getOptionalShape();
     auto dtype = ndarray.getDtype();
     os << "ndarray<";
+    if (!shape) {
+      os << "*:";
+    } else {
+      os << "[";
+      for (auto it : llvm::enumerate(*shape)) {
+        if (it.index() > 0)
+          os << ",";
+        if (it.value() < 0)
+          os << "?";
+        else
+          os << it.value();
+      }
+      os << "]:";
+    }
     if (dtype != unknownType)
       os.printType(dtype);
     else
@@ -85,19 +152,41 @@ namespace Numpy {
 namespace detail {
 
 struct NdArrayTypeStorage : public TypeStorage {
-  using KeyTy = Type;
-  NdArrayTypeStorage(Type optionalDtype) : optionalDtype(optionalDtype) {}
-  bool operator==(const KeyTy &other) const { return optionalDtype == other; }
+  using KeyTy = std::pair<Type, llvm::Optional<ArrayRef<int64_t>>>;
+  NdArrayTypeStorage(Type dtype, int rank, const int64_t *shapeElements)
+      : dtype(dtype), rank(rank), shapeElements(shapeElements) {}
+  bool operator==(const KeyTy &key) const {
+    return key == KeyTy(dtype, getOptionalShape());
+  }
   static llvm::hash_code hashKey(const KeyTy &key) {
-    return llvm::hash_combine(key);
+    if (key.second) {
+      return llvm::hash_combine(key.first, *key.second);
+    } else {
+      return llvm::hash_combine(key.first, -1);
+    }
   }
   static NdArrayTypeStorage *construct(TypeStorageAllocator &allocator,
                                        const KeyTy &key) {
+    int rank = -1;
+    const int64_t *shapeElements = nullptr;
+    if (key.second.hasValue()) {
+      auto allocElements = allocator.copyInto(*key.second);
+      rank = key.second->size();
+      shapeElements = allocElements.data();
+    }
     return new (allocator.allocate<NdArrayTypeStorage>())
-        NdArrayTypeStorage(key);
+        NdArrayTypeStorage(key.first, rank, shapeElements);
   }
 
-  Type optionalDtype;
+  llvm::Optional<ArrayRef<int64_t>> getOptionalShape() const {
+    if (rank < 0)
+      return llvm::None;
+    return ArrayRef<int64_t>(shapeElements, rank);
+  }
+
+  Type dtype;
+  int rank;
+  const int64_t *shapeElements;
 };
 
 } // namespace detail
@@ -105,16 +194,21 @@ struct NdArrayTypeStorage : public TypeStorage {
 } // namespace NPCOMP
 } // namespace mlir
 
-NdArrayType NdArrayType::get(Type dtype) {
+NdArrayType NdArrayType::get(Type dtype,
+                             llvm::Optional<ArrayRef<int64_t>> shape) {
   assert(dtype && "dtype cannot be null");
-  return Base::get(dtype.getContext(), NumpyTypes::NdArray, dtype);
+  return Base::get(dtype.getContext(), NumpyTypes::NdArray, dtype, shape);
 }
 
 bool NdArrayType::hasKnownDtype() {
   return getDtype() != Basicpy::UnknownType::get(getContext());
 }
 
-Type NdArrayType::getDtype() { return getImpl()->optionalDtype; }
+Type NdArrayType::getDtype() { return getImpl()->dtype; }
+
+llvm::Optional<ArrayRef<int64_t>> NdArrayType::getOptionalShape() {
+  return getImpl()->getOptionalShape();
+}
 
 Typing::CPA::TypeNode *
 NdArrayType::mapToCPAType(Typing::CPA::Context &context) {

pytest/NumpyCompiler/array_basics.py

@@ -13,6 +13,6 @@ global_data = (np.zeros((2, 3)) + [1.0, 2.0, 3.0] * np.reshape([1.0, 2.0],
 @import_global
 def global_array_to_const():
   # CHECK: %[[CST:.*]] =  constant dense<{{\[\[}}1.000000e+00, 2.000000e+00, 3.000000e+00], [2.000000e+00, 4.000000e+00, 6.000000e+00]]> : tensor<2x3xf64>
-  # CHECK: numpy.create_array_from_tensor %[[CST]] : (tensor<2x3xf64>) -> !numpy.ndarray<f64>
+  # CHECK: numpy.create_array_from_tensor %[[CST]] : (tensor<2x3xf64>) -> !numpy.ndarray<*:f64>
   local_data = global_data
   return local_data

pytest/NumpyCompiler/ndarray_inference.py

@@ -16,7 +16,7 @@ b = np.asarray([3.0, 4.0])
 # Test the basic flow of invoking a ufunc call with constants captured from
 # a global using explicit function syntax (np.add(a, b)).
 # CHECK-LABEL: func @global_add
-# CHECK-SAME: -> !numpy.ndarray<f64>
+# CHECK-SAME: -> !numpy.ndarray<*:f64>
 @import_global
 def global_add():
   # CHECK-NOT: UnknownType

pytest/NumpyCompiler/ufunc.py

@@ -25,7 +25,7 @@ def global_add():
   # CHECK-DAG: %[[A:.*]] = numpy.copy_to_tensor %[[A_ARRAY]]
   # CHECK-DAG: %[[B:.*]] = numpy.copy_to_tensor %[[B_ARRAY]]
   # CHECK: %[[R_TENSOR:.*]] = numpy.builtin_ufunc_call<"numpy.add"> (%[[A]], %[[B]]) : (tensor<*xf64>, tensor<*xf64>) -> tensor<*x!basicpy.UnknownType>
-  # CHECK: numpy.create_array_from_tensor %[[R_TENSOR]] : (tensor<*x!basicpy.UnknownType>) -> !numpy.ndarray<?>
+  # CHECK: numpy.create_array_from_tensor %[[R_TENSOR]] : (tensor<*x!basicpy.UnknownType>) -> !numpy.ndarray<*:?>
   return np.add(a, b)
 
 

test/Dialect/Numpy/types_attrs.mlir

@@ -1,13 +1,21 @@
 // RUN: npcomp-opt -split-input-file %s | npcomp-opt | FileCheck --dump-input=fail %s
 
 // CHECK-LABEL: @ndarray_no_dtype
-// CHECK: !numpy.ndarray<?>
-func @ndarray_no_dtype(%arg0 : !numpy.ndarray<?>) -> !numpy.ndarray<?> {
-  return %arg0 : !numpy.ndarray<?>
+// CHECK: !numpy.ndarray<*:?>
+func @ndarray_no_dtype(%arg0 : !numpy.ndarray<*:?>) -> !numpy.ndarray<*:?> {
+  return %arg0 : !numpy.ndarray<*:?>
 }
 
+// -----
 // CHECK-LABEL: @ndarray_dtype
-// CHECK: !numpy.ndarray<i32>
-func @ndarray_dtype(%arg0 : !numpy.ndarray<i32>) -> !numpy.ndarray<i32> {
-  return %arg0 : !numpy.ndarray<i32>
+// CHECK: !numpy.ndarray<*:i32>
+func @ndarray_dtype(%arg0 : !numpy.ndarray<*:i32>) -> !numpy.ndarray<*:i32> {
+  return %arg0 : !numpy.ndarray<*:i32>
+}
+
+// -----
+// CHECK-LABEL: @ndarray_ranked
+// CHECK: !numpy.ndarray<[1,?,3]:i32>
+func @ndarray_ranked(%arg0 : !numpy.ndarray<[1,?,3]:i32>) -> !numpy.ndarray<[1,?,3]:i32> {
+  return %arg0 : !numpy.ndarray<[1,?,3]:i32>
 }