@datarecce/ui 0.1.29 → 0.1.31

package/recce-source/js/src/lib/dataGrid/propertyBased.test.ts

@@ -0,0 +1,858 @@
+/**
+ * @file propertyBased.test.ts
+ * @description Property-based tests for data grid functions using fast-check 4.4.0
+ *
+ * These tests verify invariants that should hold for ANY valid input,
+ * automatically generating test cases and shrinking failures to minimal examples.
+ *
+ * Key invariants tested:
+ * - Structural consistency: output rows have consistent shape
+ * - Row count preservation: no data loss or duplication
+ * - Primary key uniqueness: PK values are unique in output
+ * - Diff symmetry: identical inputs produce no modifications
+ * - Column alignment: output columns match input schema
+ */
+
+import type { ColDef, ColGroupDef } from "ag-grid-community";
+import fc from "fast-check";
+import {
+  ColumnRenderMode,
+  ColumnType,
+  DataFrame,
+  RowObjectType,
+} from "@/lib/api/types";
+import { toDataDiffGrid } from "@/lib/dataGrid/generators/toDataDiffGrid";
+import { toDataGrid } from "@/lib/dataGrid/generators/toDataGrid";
+import { toValueDiffGrid } from "@/lib/dataGrid/generators/toValueDiffGrid";
+
+// ============================================================================
+// Types
+// ============================================================================
+
+/**
+ * Extended column type returned by grid functions (AG Grid)
+ */
+type ExtendedColumn = (ColDef<RowObjectType> | ColGroupDef<RowObjectType>) & {
+  columnType?: ColumnType;
+  columnRenderMode?: ColumnRenderMode;
+};
+
+// ============================================================================
+// Mocks
+// ============================================================================
+
+jest.mock("ag-grid-community", () => ({
+  ModuleRegistry: { registerModules: jest.fn() },
+  AllCommunityModule: {},
+}));
+
+jest.mock("@/components/ui/mui", () => ({
+  Box: ({ children }: { children: React.ReactNode }) => children,
+  Flex: ({ children }: { children: React.ReactNode }) => children,
+  Icon: () => null,
+  IconButton: () => null,
+  Menu: {
+    Root: ({ children }: { children: React.ReactNode }) => children,
+    Trigger: ({ children }: { children: React.ReactNode }) => children,
+    Content: ({ children }: { children: React.ReactNode }) => children,
+    Item: ({ children }: { children: React.ReactNode }) => children,
+    ItemGroup: ({ children }: { children: React.ReactNode }) => children,
+    Positioner: ({ children }: { children: React.ReactNode }) => children,
+  },
+  Portal: ({ children }: { children: React.ReactNode }) => children,
+  Text: ({ children }: { children: React.ReactNode }) => children,
+}));
+
+// ============================================================================
+// Custom Arbitraries for Data Grid Types
+// ============================================================================
+
+/**
+ * Arbitrary for ColumnType enum values
+ */
+const columnTypeArb: fc.Arbitrary<ColumnType> = fc.constantFrom(
+  "number",
+  "integer",
+  "text",
+  "boolean",
+  "date",
+  "datetime",
+  "timedelta",
+  "unknown",
+);
+
+/**
+ * Arbitrary for column names - valid identifiers without special chars
+ */
+const columnNameArb: fc.Arbitrary<string> = fc
+  .stringMatching(/^[a-z][a-z0-9_]{0,19}$/)
+  .filter((s) => s.length > 0 && s !== "id");
+
+/**
+ * Arbitrary for cell values based on column type
+ * Excludes NaN/Infinity by default for cleaner property tests
+ */
+function cellValueArb(colType: ColumnType): fc.Arbitrary<unknown> {
+  switch (colType) {
+    case "integer":
+      return fc.oneof(
+        fc.integer({ min: -1000000, max: 1000000 }),
+        fc.constant(null),
+      );
+    case "number":
+      return fc.oneof(
+        fc.double({
+          noNaN: true,
+          noDefaultInfinity: true,
+          min: -1e6,
+          max: 1e6,
+        }),
+        fc.constant(null),
+      );
+    case "text":
+      return fc.oneof(
+        fc.string({ minLength: 0, maxLength: 50 }),
+        fc.constant(null),
+      );
+    case "boolean":
+      return fc.oneof(fc.boolean(), fc.constant(null));
+    case "date":
+      return fc.oneof(
+        fc
+          .date({ noInvalidDate: true })
+          .map((d) => d.toISOString().split("T")[0]),
+        fc.constant(null),
+      );
+    case "datetime":
+      return fc.oneof(
+        fc.date({ noInvalidDate: true }).map((d) => d.toISOString()),
+        fc.constant(null),
+      );
+    case "timedelta":
+      return fc.oneof(
+        fc
+          .tuple(fc.nat({ max: 23 }), fc.nat({ max: 59 }), fc.nat({ max: 59 }))
+          .map(
+            ([h, m, s]) =>
+              `${h}:${String(m).padStart(2, "0")}:${String(s).padStart(2, "0")}`,
+          ),
+        fc.constant(null),
+      );
+    case "unknown":
+    default:
+      return fc.oneof(
+        fc.string({ maxLength: 20 }),
+        fc.integer(),
+        fc.boolean(),
+        fc.constant(null),
+      );
+  }
+}
+
+/**
+ * Arbitrary for unique primary key values (integers)
+ */
+const primaryKeyValueArb: fc.Arbitrary<number> = fc.integer({
+  min: 1,
+  max: 100000,
+});
+
+/**
+ * Configuration for DataFrame generation
+ */
+interface DataFrameConfig {
+  minColumns?: number;
+  maxColumns?: number;
+  minRows?: number;
+  maxRows?: number;
+  includePrimaryKey?: boolean;
+}
+
+/**
+ * Arbitrary for a complete DataFrame with consistent structure
+ *
+ * Generates columns with types, then generates row data that matches
+ * the column types. Optionally includes a guaranteed integer primary key.
+ */
+function dataFrameArb(config: DataFrameConfig = {}): fc.Arbitrary<DataFrame> {
+  const {
+    minColumns = 1,
+    maxColumns = 5,
+    minRows = 0,
+    maxRows = 20,
+    includePrimaryKey = true,
+  } = config;
+
+  return fc
+    .tuple(
+      // Generate unique column names
+      fc.uniqueArray(columnNameArb, {
+        minLength: minColumns,
+        maxLength: maxColumns,
+      }),
+      // Generate column types for each column
+      fc.array(columnTypeArb, { minLength: minColumns, maxLength: maxColumns }),
+      // Number of rows
+      fc.nat({ max: maxRows - minRows }),
+    )
+    .chain(([names, types, extraRows]) => {
+      // Ensure we have matching lengths
+      const columnCount = Math.min(names.length, types.length);
+      const trimmedNames = names.slice(0, columnCount);
+      const trimmedTypes = types.slice(0, columnCount);
+      const rowCount = minRows + extraRows;
+
+      // Build column definitions
+      // Filter out "id" from generated names when includePrimaryKey is true
+      // to avoid duplicate column keys
+      const filteredNames = includePrimaryKey
+        ? trimmedNames.filter((name) => name !== "id")
+        : trimmedNames;
+      const filteredTypes = includePrimaryKey
+        ? trimmedTypes.filter((_, i) => trimmedNames[i] !== "id")
+        : trimmedTypes;
+
+      const columns: DataFrame["columns"] = includePrimaryKey
+        ? [
+            { key: "id", name: "id", type: "integer" },
+            ...filteredNames.map((name, i) => ({
+              key: name,
+              name: name,
+              type: filteredTypes[i],
+            })),
+          ]
+        : filteredNames.map((name, i) => ({
+            key: name,
+            name: name,
+            type: filteredTypes[i],
+          }));
+
+      // Generate row data matching column types
+      // Primary key uses dedicated arbitrary (non-null, unique), other columns use type-based
+      const rowArb = includePrimaryKey
+        ? fc.tuple(primaryKeyValueArb, ...filteredTypes.map(cellValueArb))
+        : fc.tuple(...filteredTypes.map(cellValueArb));
+
+      return fc
+        .uniqueArray(rowArb, {
+          minLength: rowCount,
+          maxLength: rowCount,
+          comparator: (a, b) => {
+            // Ensure unique primary keys when included
+            if (includePrimaryKey) {
+              return a[0] === b[0];
+            }
+            return false;
+          },
+        })
+        .map((data) => ({
+          columns,
+          data: data as DataFrame["data"],
+        }));
+    });
+}
+
+/**
+ * Arbitrary for in_a/in_b flag pairs
+ * At least one must be true (rows with both false are invalid - not in either dataset)
+ */
+const validInFlagsArb: fc.Arbitrary<[boolean, boolean]> = fc.constantFrom(
+  [true, true] as [boolean, boolean], // In both (unchanged or modified)
+  [true, false] as [boolean, boolean], // In base only (removed)
+  [false, true] as [boolean, boolean], // In current only (added)
+);
+
+/**
+ * Arbitrary for joined DataFrame (used by toValueDiffGrid)
+ *
+ * The input format for toValueDiffGrid is:
+ * - Columns: [pk, ...valueColumns, in_a, in_b]
+ * - Data: [pkValue, ...values, inA, inB]
+ *
+ * For modified rows, there are TWO input rows with the same PK:
+ * - One with in_a=true, in_b=false (base version)
+ * - One with in_a=false, in_b=true (current version)
+ *
+ * The function then merges these into base__/current__ prefixed output columns.
+ */
+function joinedDataFrameArb(config: DataFrameConfig = {}): fc.Arbitrary<{
+  dataframe: DataFrame;
+  primaryKeys: string[];
+}> {
+  const { minRows = 1, maxRows = 20 } = config;
+
+  return fc
+    .tuple(
+      // Generate 1-3 value column names
+      fc.uniqueArray(columnNameArb, { minLength: 1, maxLength: 3 }),
+      // Generate column types
+      fc.array(columnTypeArb, { minLength: 1, maxLength: 3 }),
+      // Number of unique PKs
+      fc.nat({ max: Math.max(0, maxRows - minRows) }),
+    )
+    .chain(([names, types, extraRows]) => {
+      const columnCount = Math.min(names.length, types.length);
+      const trimmedNames = names.slice(0, columnCount);
+      const trimmedTypes = types.slice(0, columnCount);
+      const pkCount = minRows + extraRows;
+
+      // Build column definitions: [id, ...valueColumns, in_a, in_b]
+      const columns: DataFrame["columns"] = [
+        { key: "id", name: "id", type: "integer" },
+        ...trimmedNames.map((name, i) => ({
+          key: name,
+          name: name,
+          type: trimmedTypes[i],
+        })),
+        { key: "in_a", name: "in_a", type: "boolean" },
+        { key: "in_b", name: "in_b", type: "boolean" },
+      ];
+
+      // Generate unique PK values
+      return fc
+        .uniqueArray(primaryKeyValueArb, {
+          minLength: pkCount,
+          maxLength: pkCount,
+        })
+        .chain((pkValues) => {
+          // For each PK, generate status and values
+          const rowGenArb = fc.tuple(
+            validInFlagsArb,
+            fc.tuple(...trimmedTypes.map(cellValueArb)),
+          );
+
+          return fc
+            .array(rowGenArb, { minLength: pkCount, maxLength: pkCount })
+            .map((rowSpecs) => {
+              const data: unknown[][] = [];
+
+              pkValues.forEach((pk, i) => {
+                const [[inA, inB], values] = rowSpecs[i];
+
+                if (inA && inB) {
+                  // Row in both - single row with in_a=true, in_b=true
+                  data.push([pk, ...values, true, true]);
+                } else if (inA && !inB) {
+                  // Removed - single row with in_a=true, in_b=false
+                  data.push([pk, ...values, true, false]);
+                } else if (!inA && inB) {
+                  // Added - single row with in_a=false, in_b=true
+                  data.push([pk, ...values, false, true]);
+                }
+              });
+
+              return {
+                dataframe: {
+                  columns,
+                  data: data as DataFrame["data"],
+                },
+                primaryKeys: ["id"],
+              };
+            });
+        });
+    });
+}
+
+// ============================================================================
+// Helper Functions
+// ============================================================================
+
+/**
+ * Extract column field from a column or column group
+ * Fields are lowercased for consistent comparison
+ */
+function getColumnKey(col: ExtendedColumn): string | undefined {
+  if ("field" in col && col.field) {
+    return col.field.toLowerCase();
+  }
+  if ("children" in col && Array.isArray(col.children) && col.children[0]) {
+    const firstChild = col.children[0] as ColDef<RowObjectType>;
+    if ("field" in firstChild && firstChild.field) {
+      // Extract base column name from "base__colname"
+      const field = firstChild.field;
+      if (field.startsWith("base__")) {
+        return field.replace("base__", "").toLowerCase();
+      }
+      return field.toLowerCase();
+    }
+  }
+  return undefined;
+}
+
+/**
+ * Extract all column keys from grid result, handling column groups
+ * Keys are lowercased for consistent comparison
+ */
+function getResultColumnKeys(columns: ExtendedColumn[]): string[] {
+  return columns.map(getColumnKey).filter((k): k is string => k !== undefined);
+}
+
+/**
+ * Check if all rows have consistent keys
+ */
+function rowsHaveConsistentKeys(rows: RowObjectType[]): boolean {
+  if (rows.length === 0) return true;
+
+  const referenceKeys = new Set(
+    Object.keys(rows[0]).filter((k) => !k.startsWith("__") && k !== "_index"),
+  );
+
+  return rows.every((row) => {
+    const rowKeys = new Set(
+      Object.keys(row).filter((k) => !k.startsWith("__") && k !== "_index"),
+    );
+    if (rowKeys.size !== referenceKeys.size) return false;
+    for (const key of referenceKeys) {
+      if (!rowKeys.has(key)) return false;
+    }
+    return true;
+  });
+}
+
+// ============================================================================
+// Property-Based Tests: toDataGrid
+// ============================================================================
+
+describe("Property-based tests: toDataGrid", () => {
+  test("produces correct row count for any valid DataFrame", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 0, maxRows: 50 }), (df) => {
+        const result = toDataGrid(df, { primaryKeys: ["id"] });
+        return result.rows.length === df.data.length;
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("all rows have consistent structure", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 1, maxRows: 30 }), (df) => {
+        const result = toDataGrid(df, { primaryKeys: ["id"] });
+        return rowsHaveConsistentKeys(result.rows);
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("row data contains all column keys from input", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 1, maxRows: 20 }), (df) => {
+        const result = toDataGrid(df, { primaryKeys: ["id"] });
+        const inputKeys = new Set(df.columns.map((c) => c.key.toLowerCase()));
+
+        // Every row should have all input column keys
+        return result.rows.every((row) => {
+          for (const key of inputKeys) {
+            if (!(key in row)) return false;
+          }
+          return true;
+        });
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("primary key values are preserved in output rows", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 1, maxRows: 30 }), (df) => {
+        const result = toDataGrid(df, { primaryKeys: ["id"] });
+        const inputPKs = new Set(df.data.map((row) => row[0]));
+        const outputPKs = new Set(result.rows.map((row) => row.id));
+
+        // All input PKs should appear in output
+        for (const pk of inputPKs) {
+          if (!outputPKs.has(pk as number)) return false;
+        }
+        return true;
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("handles empty DataFrame without error", () => {
+    fc.assert(
+      fc.property(
+        dataFrameArb({ minRows: 0, maxRows: 0, minColumns: 1 }),
+        (df) => {
+          const result = toDataGrid(df, { primaryKeys: ["id"] });
+          return result.rows.length === 0 && result.columns.length > 0;
+        },
+      ),
+      { numRuns: 50 },
+    );
+  });
+
+  test("__status is undefined for single DataFrame (no diff)", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 1, maxRows: 20 }), (df) => {
+        const result = toDataGrid(df, { primaryKeys: ["id"] });
+        return result.rows.every((row) => row.__status === undefined);
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("output columns include all input column keys", () => {
+    fc.assert(
+      fc.property(
+        dataFrameArb({ minRows: 1, maxRows: 10, minColumns: 2 }),
+        (df) => {
+          const result = toDataGrid(df, { primaryKeys: ["id"] });
+          const resultKeys = getResultColumnKeys(result.columns);
+          const inputKeys = df.columns.map((c) => c.key.toLowerCase());
+
+          // All input column keys should appear in result columns
+          return inputKeys.every((key) => resultKeys.includes(key));
+        },
+      ),
+      { numRuns: 100 },
+    );
+  });
+});
+
+// ============================================================================
+// Property-Based Tests: toDataDiffGrid
+// ============================================================================
+
+describe("Property-based tests: toDataDiffGrid", () => {
+  test("identical DataFrames produce no modified/added/removed rows", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 1, maxRows: 30 }), (df) => {
+        const result = toDataDiffGrid(df, df, { primaryKeys: ["id"] });
+
+        // All rows should have undefined status (unchanged)
+        return result.rows.every((row) => row.__status === undefined);
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("row count with identical inputs equals input row count", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 0, maxRows: 30 }), (df) => {
+        const result = toDataDiffGrid(df, df, { primaryKeys: ["id"] });
+        return result.rows.length === df.data.length;
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("empty base produces all 'added' rows", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 1, maxRows: 20 }), (df) => {
+        const emptyBase: DataFrame = { columns: df.columns, data: [] };
+        const result = toDataDiffGrid(emptyBase, df, { primaryKeys: ["id"] });
+
+        return (
+          result.rows.length === df.data.length &&
+          result.rows.every((row) => row.__status === "added")
+        );
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("empty current produces all 'removed' rows", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 1, maxRows: 20 }), (df) => {
+        const emptyCurrent: DataFrame = { columns: df.columns, data: [] };
+        const result = toDataDiffGrid(df, emptyCurrent, {
+          primaryKeys: ["id"],
+        });
+
+        return (
+          result.rows.length === df.data.length &&
+          result.rows.every((row) => row.__status === "removed")
+        );
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("all rows have both base__ and current__ prefixed columns", () => {
+    fc.assert(
+      fc.property(
+        dataFrameArb({ minRows: 1, maxRows: 20, minColumns: 2 }),
+        (df) => {
+          const result = toDataDiffGrid(df, df, {
+            primaryKeys: ["id"],
+            displayMode: "side_by_side",
+          });
+
+          // For side_by_side mode, non-PK columns should have base__/current__ versions
+          const valueColumns = df.columns.slice(1); // Skip 'id'
+          return result.rows.every((row) =>
+            valueColumns.every((col) => {
+              const baseKey = `base__${col.key}`.toLowerCase();
+              const currentKey = `current__${col.key}`.toLowerCase();
+              return baseKey in row && currentKey in row;
+            }),
+          );
+        },
+      ),
+      { numRuns: 100 },
+    );
+  });
+
+  test("changedOnly=true with identical inputs returns empty rows", () => {
+    fc.assert(
+      fc.property(dataFrameArb({ minRows: 1, maxRows: 20 }), (df) => {
+        const result = toDataDiffGrid(df, df, {
+          primaryKeys: ["id"],
+          changedOnly: true,
+        });
+
+        // No changes means no rows when changedOnly is true
+        return result.rows.length === 0;
+      }),
+      { numRuns: 100 },
+    );
+  });
+
+  test("total rows = base-only + current-only + matched", () => {
+    fc.assert(
+      fc.property(
+        dataFrameArb({ minRows: 1, maxRows: 15 }),
+        dataFrameArb({ minRows: 1, maxRows: 15 }),
+        (base, current) => {
+          const result = toDataDiffGrid(base, current, { primaryKeys: ["id"] });
+
+          const basePKs = new Set(base.data.map((r) => r[0]));
+          const currentPKs = new Set(current.data.map((r) => r[0]));
+
+          // Count expected rows
+          const baseOnly = [...basePKs].filter(
+            (pk) => !currentPKs.has(pk),
+          ).length;
+          const currentOnly = [...currentPKs].filter(
+            (pk) => !basePKs.has(pk),
+          ).length;
+          const matched = [...basePKs].filter((pk) =>
+            currentPKs.has(pk),
+          ).length;
+
+          const expectedTotal = baseOnly + currentOnly + matched;
+          return result.rows.length === expectedTotal;
+        },
+      ),
+      { numRuns: 100 },
+    );
+  });
+});
+
+// ============================================================================
+// Property-Based Tests: toValueDiffGrid
+// ============================================================================
+
+describe("Property-based tests: toValueDiffGrid", () => {
+  test("row count matches input DataFrame row count", () => {
+    fc.assert(
+      fc.property(
+        joinedDataFrameArb({ minRows: 1, maxRows: 30 }),
+        ({ dataframe, primaryKeys }) => {
+          const result = toValueDiffGrid(dataframe, primaryKeys, {});
+          return result.rows.length === dataframe.data.length;
+        },
+      ),
+      { numRuns: 100 },
+    );
+  });
+
+  test("primary key values are unique in output", () => {
+    fc.assert(
+      fc.property(
+        joinedDataFrameArb({ minRows: 1, maxRows: 30 }),
+        ({ dataframe, primaryKeys }) => {
+          const result = toValueDiffGrid(dataframe, primaryKeys, {});
+          const pkValues = result.rows.map((row) => row[primaryKeys[0]]);
+          const uniquePKs = new Set(pkValues);
+          return uniquePKs.size === pkValues.length;
+        },
+      ),
+      { numRuns: 100 },
+    );
+  });
+});
+
+// ============================================================================
+// Property-Based Tests: Edge Cases with Special Values
+// ============================================================================
+
+describe("Property-based tests: special values", () => {
+  /**
+   * Arbitrary that includes NaN and Infinity for stress testing
+   */
+  const numericWithSpecialsArb = fc.oneof(
+    fc.double({ noNaN: true, noDefaultInfinity: true }),
+    fc.constant(NaN),
+    fc.constant(Infinity),
+    fc.constant(-Infinity),
+    fc.constant(null),
+  );
+
+  test("toDataGrid handles any numeric value without throwing", () => {
+    fc.assert(
+      fc.property(
+        fc.array(numericWithSpecialsArb, { minLength: 1, maxLength: 20 }),
+        (values) => {
+          const df: DataFrame = {
+            columns: [
+              { key: "id", name: "id", type: "integer" },
+              { key: "value", name: "value", type: "number" },
+            ],
+            data: values.map((v, i) => [i + 1, v]),
+          };
+
+          // Should not throw
+          const result = toDataGrid(df, { primaryKeys: ["id"] });
+          return result.rows.length === values.length;
+        },
+      ),
+      { numRuns: 100 },
+    );
+  });
+
+  test("toDataDiffGrid handles NaN comparison correctly", () => {
+    fc.assert(
+      fc.property(fc.nat({ max: 100 }), (seed) => {
+        // NaN === NaN should be true via lodash.isEqual
+        const df: DataFrame = {
+          columns: [
+            { key: "id", name: "id", type: "integer" },
+            { key: "value", name: "value", type: "number" },
+          ],
+          data: [[seed, NaN]],
+        };
+
+        const result = toDataDiffGrid(df, df, { primaryKeys: ["id"] });
+
+        // NaN in both base and current should NOT be marked as modified
+        // (lodash.isEqual treats NaN === NaN as true)
+        return result.rows.every((row) => row.__status === undefined);
+      }),
+      { numRuns: 50 },
+    );
+  });
+
+  test("null vs undefined handling in comparisons", () => {
+    fc.assert(
+      fc.property(fc.nat({ max: 100 }), (id) => {
+        const base: DataFrame = {
+          columns: [
+            { key: "id", name: "id", type: "integer" },
+            { key: "value", name: "value", type: "text" },
+          ],
+          data: [[id, null]],
+        };
+        const current: DataFrame = {
+          columns: [
+            { key: "id", name: "id", type: "integer" },
+            { key: "value", name: "value", type: "text" },
+          ],
+          data: [[id, undefined]],
+        };
+
+        // Should handle both without throwing
+        const result = toDataDiffGrid(base, current, { primaryKeys: ["id"] });
+        return result.rows.length === 1;
+      }),
+      { numRuns: 50 },
+    );
+  });
+});
+
+// ============================================================================
+// Property-Based Tests: Stress Testing with Large Data
+// ============================================================================
+
+describe("Property-based tests: stress testing", () => {
+  test("handles wide DataFrames (many columns)", () => {
+    fc.assert(
+      fc.property(
+        dataFrameArb({
+          minColumns: 10,
+          maxColumns: 20,
+          minRows: 5,
+          maxRows: 10,
+        }),
+        (df) => {
+          const result = toDataGrid(df, { primaryKeys: ["id"] });
+          return result.rows.length === df.data.length;
+        },
+      ),
+      { numRuns: 50 },
+    );
+  });
+
+  test("handles tall DataFrames (many rows)", () => {
+    fc.assert(
+      fc.property(
+        dataFrameArb({
+          minColumns: 2,
+          maxColumns: 3,
+          minRows: 50,
+          maxRows: 100,
+        }),
+        (df) => {
+          const result = toDataGrid(df, { primaryKeys: ["id"] });
+          return result.rows.length === df.data.length;
+        },
+      ),
+      { numRuns: 30 },
+    );
+  });
+
+  test("diff grid handles large mismatched DataFrames", () => {
+    fc.assert(
+      fc.property(
+        // Generate a shared column schema first, then two DataFrames with same schema
+        fc
+          .tuple(
+            fc.uniqueArray(columnNameArb, { minLength: 2, maxLength: 4 }),
+            fc.array(columnTypeArb, { minLength: 2, maxLength: 4 }),
+          )
+          .chain(([names, types]) => {
+            const columnCount = Math.min(names.length, types.length);
+            const columns: DataFrame["columns"] = [
+              { key: "id", name: "id", type: "integer" },
+              ...names.slice(0, columnCount).map((name, i) => ({
+                key: name,
+                name: name,
+                type: types[i],
+              })),
+            ];
+
+            // Generate two DataFrames with the same schema but different data
+            const rowArb = fc.tuple(
+              primaryKeyValueArb,
+              ...types.slice(0, columnCount).map(cellValueArb),
+            );
+
+            return fc.tuple(
+              fc
+                .uniqueArray(rowArb, {
+                  minLength: 20,
+                  maxLength: 40,
+                  comparator: (a, b) => a[0] === b[0],
+                })
+                .map((data) => ({ columns, data: data as DataFrame["data"] })),
+              fc
+                .uniqueArray(rowArb, {
+                  minLength: 20,
+                  maxLength: 40,
+                  comparator: (a, b) => a[0] === b[0],
+                })
+                .map((data) => ({ columns, data: data as DataFrame["data"] })),
+            );
+          }),
+        ([base, current]) => {
+          const result = toDataDiffGrid(base, current, { primaryKeys: ["id"] });
+
+          // Should complete without error and produce reasonable output
+          return result.rows.length >= 0 && result.columns.length > 0;
+        },
+      ),
+      { numRuns: 30 },
+    );
+  });
+});