Merge #26452

26452: importccl: make spans overlap during IMPORT with transform r=mjibson a=mjibson

Force the backup descriptor spans to overlap correctly. This prevents
various extra ranges from being created during RESTORE of data created
with IMPORT with transform. The IMPORT commands will need to be re-run
to generate correct spans in the descriptors.

Fixes #26375

Release note: None

Co-authored-by: Matt Jibson <[email protected]>

pkg/ccl/importccl/import_stmt.go

@@ -602,6 +602,23 @@ func doDistributedCSVTransform(
 		})
 	}
 
+	// The returned spans are from the SSTs themselves, and so don't perfectly
+	// overlap. Sort the files so we can fix the spans to be correctly
+	// overlapping. This is needed because RESTORE splits at both the start
+	// and end of each SST, and so there are tiny ranges (like {NULL-/0/0} at
+	// the start) that get created. During non-transform IMPORT this isn't a
+	// problem because it only splits on the end key. Replicate that behavior
+	// here by copying the end key from each span to the start key of the next.
+	sort.Slice(backupDesc.Files, func(i, j int) bool {
+		return backupDesc.Files[i].Span.Key.Compare(backupDesc.Files[j].Span.Key) < 0
+	})
+	tableSpan := tableDesc.TableSpan()
+	backupDesc.Files[0].Span.Key = tableSpan.Key
+	for i := 1; i < len(backupDesc.Files); i++ {
+		backupDesc.Files[i].Span.Key = backupDesc.Files[i-1].Span.EndKey
+	}
+	backupDesc.Files[len(backupDesc.Files)-1].Span.EndKey = tableSpan.EndKey
+
 	dest, err := storageccl.ExportStorageConfFromURI(transformOnly)
 	if err != nil {
 		return err

pkg/ccl/importccl/import_stmt_test.go

@@ -656,6 +656,15 @@ func TestImportCSVStmt(t *testing.T) {
 					t.Fatal("expected > 1 SST files")
 				}
 			}
+
+			// Verify spans don't have trailing '/0'.
+			ranges := sqlDB.QueryStr(t, `SHOW testing_ranges FROM TABLE t`)
+			for _, r := range ranges {
+				const end = `/0`
+				if strings.HasSuffix(r[0], end) || strings.HasSuffix(r[1], end) {
+					t.Errorf("bad span: %s - %s", r[0], r[1])
+				}
+			}
 		})
 	}
 

pkg/ccl/importccl/sst_writer_proc.go

@@ -167,12 +167,12 @@ func (sp *sstWriter) Run(ctx context.Context, wg *sync.WaitGroup) {
 					if chunk > 0 {
 						name = fmt.Sprintf("%d-%s", chunk, name)
 					}
+					end := span.End
+					if sst.more {
+						end = sst.span.EndKey
+					}
 
 					if sp.spec.Destination == "" {
-						end := span.End
-						if sst.more {
-							end = sst.span.EndKey
-						}
 						if err := sp.db.AdminSplit(ctx, end, end); err != nil {
 							return err
 						}
@@ -231,7 +231,7 @@ func (sp *sstWriter) Run(ctx context.Context, wg *sync.WaitGroup) {
 						),
 						sqlbase.DatumToEncDatum(
 							sqlbase.ColumnType{SemanticType: sqlbase.ColumnType_BYTES},
-							tree.NewDBytes(tree.DBytes(sst.span.EndKey)),
+							tree.NewDBytes(tree.DBytes(end)),
 						),
 					}
 					cs, err := sp.out.EmitRow(ctx, row)