[Mlir-commits] [mlir] [mlir][sparse] Migration to sparse_tensor.print (PR #83377)

[Aart Bik]

================
@@ -140,33 +140,94 @@ module {
     %b4 = sparse_tensor.convert %sb : tensor<8x4xf64> to tensor<8x4xf64, #DCSR>
 
     //
-    // Sanity check on stored entries before going into the computations.
-    //
-    // CHECK:      32
-    // CHECK-NEXT: 32
-    // CHECK-NEXT: 4
-    // CHECK-NEXT: 4
-    // CHECK-NEXT: 32
-    // CHECK-NEXT: 32
-    // CHECK-NEXT: 8
-    // CHECK-NEXT: 8
-    //
-    %noea1 = sparse_tensor.number_of_entries %a1 : tensor<4x8xf64, #CSR>
-    %noea2 = sparse_tensor.number_of_entries %a2 : tensor<4x8xf64, #DCSR>
-    %noea3 = sparse_tensor.number_of_entries %a3 : tensor<4x8xf64, #CSR>
-    %noea4 = sparse_tensor.number_of_entries %a4 : tensor<4x8xf64, #DCSR>
-    %noeb1 = sparse_tensor.number_of_entries %b1 : tensor<8x4xf64, #CSR>
-    %noeb2 = sparse_tensor.number_of_entries %b2 : tensor<8x4xf64, #DCSR>
-    %noeb3 = sparse_tensor.number_of_entries %b3 : tensor<8x4xf64, #CSR>
-    %noeb4 = sparse_tensor.number_of_entries %b4 : tensor<8x4xf64, #DCSR>
-    vector.print %noea1 : index
-    vector.print %noea2 : index
-    vector.print %noea3 : index
-    vector.print %noea4 : index
-    vector.print %noeb1 : index
-    vector.print %noeb2 : index
-    vector.print %noeb3 : index
-    vector.print %noeb4 : index
+    // Sanity check before going into the computations.
+    //
+    // CHECK:      ---- Sparse Tensor ----
+    // CHECK-NEXT: nse = 32
+    // CHECK-NEXT: pos[1] : ( 0, 8, 16, 24, 32
+    // CHECK-NEXT: crd[1] : ( 0, 1, 2, 3, 4, 5, 6, 7, 0, 1, 2, 3, 4, 5, 6, 7, 0, 1, 2, 3, 4, 5, 6, 7, 0, 1, 2, 3, 4, 5, 6, 7
+    // CHECK-NEXT: values : ( 1.1, 2.1, 3.1, 4.1, 5.1, 6.1, 7.1, 8.1, 1.2, 2.2, 3.2, 4.2, 5.2, 6.2, 7.2, 8.2, 1.3, 2.3, 3.3, 4.3, 5.3, 6.3, 7.3, 8.3, 1.4, 2.4, 3.4, 4.4, 5.4, 6.4, 7.4, 8.4
+    // CHECK-NEXT: ----
+    //
+    sparse_tensor.print %a1 : tensor<4x8xf64, #CSR>
+
+    //
+    // CHECK:      ---- Sparse Tensor ----
----------------
aartbik wrote:

note that we can make anything other than the first CHECK-NEXT

(although your approach would make it easier to add/delete examples later, so it is a "soft suggestion")

https://github.com/llvm/llvm-project/pull/83377