Fix: Use lazy encoding for utf-8 encoded string comparison

dev/src/order.ts

@@ -248,19 +248,38 @@ function compareVectors(left: ApiMapValue, right: ApiMapValue): number {
   return compareArrays(leftArray, rightArray);
 }
 
-function stringToUtf8Bytes(str: string): Uint8Array {
-  return new TextEncoder().encode(str);
-}
-
 /*!
  * Compare strings in UTF-8 encoded byte order
  * @private
  * @internal
  */
 export function compareUtf8Strings(left: string, right: string): number {
-  const leftBytes = stringToUtf8Bytes(left);
-  const rightBytes = stringToUtf8Bytes(right);
-  return compareBlobs(Buffer.from(leftBytes), Buffer.from(rightBytes));
+  let i = 0;
+  while (i < left.length && i < right.length) {
+    const leftCodePoint = left.codePointAt(i)!;
+    const rightCodePoint = right.codePointAt(i)!;
+
+    if (leftCodePoint !== rightCodePoint) {
+      if (leftCodePoint < 128 && rightCodePoint < 128) {
+        // ASCII comparison
+        return primitiveComparator(leftCodePoint, rightCodePoint);
+      } else {
+        // Lazy instantiate TextEncoder
+        const encoder = new TextEncoder();
+
+        // UTF-8 encoded byte comparison, substring 2 indexes to cover surrogate pairs
+        const leftBytes = encoder.encode(left.substring(i, i + 2));
+        const rightBytes = encoder.encode(right.substring(i, i + 2));
+        return compareBlobs(Buffer.from(leftBytes), Buffer.from(rightBytes));
+      }
+    }
+
+    // Increment by 2 for surrogate pairs, 1 otherwise
+    i += leftCodePoint > 0xffff ? 2 : 1;
+  }
+
+  // Compare lengths if all characters are equal
+  return primitiveComparator(left.length, right.length);
 }
 
 /*!

dev/system-test/firestore.ts

@@ -4086,6 +4086,20 @@ describe('Query class', () => {
     });
 
     describe('sort unicode strings', () => {
+      const expectedDocs = [
+        'b',
+        'a',
+        'h',
+        'i',
+        'c',
+        'f',
+        'e',
+        'd',
+        'g',
+        'k',
+        'j',
+      ];
+
       it('snapshot listener sorts unicode strings same as server', async () => {
         const collection = await testCollectionWithDocs({
           a: {value: 'Łukasiewicz'},
@@ -4095,10 +4109,13 @@ describe('Query class', () => {
           e: {value: 'Ｐ'},
           f: {value: '︒'},
           g: {value: '🐵'},
+          h: {value: '你好'},
+          i: {value: '你顥'},
+          j: {value: '😁'},
+          k: {value: '😀'},
         });
 
         const query = collection.orderBy('value');
-        const expectedDocs = ['b', 'a', 'c', 'f', 'e', 'd', 'g'];
 
         const getSnapshot = await query.get();
         expect(getSnapshot.docs.map(d => d.id)).to.deep.equal(expectedDocs);
@@ -4123,10 +4140,13 @@ describe('Query class', () => {
           e: {value: ['Ｐ']},
           f: {value: ['︒']},
           g: {value: ['🐵']},
+          h: {value: ['你好']},
+          i: {value: ['你顥']},
+          j: {value: ['😁']},
+          k: {value: ['😀']},
         });
 
         const query = collection.orderBy('value');
-        const expectedDocs = ['b', 'a', 'c', 'f', 'e', 'd', 'g'];
 
         const getSnapshot = await query.get();
         expect(getSnapshot.docs.map(d => d.id)).to.deep.equal(expectedDocs);
@@ -4151,10 +4171,13 @@ describe('Query class', () => {
           e: {value: {foo: 'Ｐ'}},
           f: {value: {foo: '︒'}},
           g: {value: {foo: '🐵'}},
+          h: {value: {foo: '你好'}},
+          i: {value: {foo: '你顥'}},
+          j: {value: {foo: '😁'}},
+          k: {value: {foo: '😀'}},
         });
 
         const query = collection.orderBy('value');
-        const expectedDocs = ['b', 'a', 'c', 'f', 'e', 'd', 'g'];
 
         const getSnapshot = await query.get();
         expect(getSnapshot.docs.map(d => d.id)).to.deep.equal(expectedDocs);
@@ -4179,10 +4202,13 @@ describe('Query class', () => {
           e: {value: {Ｐ: true}},
           f: {value: {'︒': true}},
           g: {value: {'🐵': true}},
+          h: {value: {你好: true}},
+          i: {value: {你顥: true}},
+          j: {value: {'😁': true}},
+          k: {value: {'😀': true}},
         });
 
         const query = collection.orderBy('value');
-        const expectedDocs = ['b', 'a', 'c', 'f', 'e', 'd', 'g'];
 
         const getSnapshot = await query.get();
         expect(getSnapshot.docs.map(d => d.id)).to.deep.equal(expectedDocs);
@@ -4207,17 +4233,25 @@ describe('Query class', () => {
           Ｐ: {value: true},
           '︒': {value: true},
           '🐵': {value: true},
+          你好: {value: true},
+          你顥: {value: true},
+          '😁': {value: true},
+          '😀': {value: true},
         });
 
         const query = collection.orderBy(FieldPath.documentId());
         const expectedDocs = [
           'Sierpiński',
           'Łukasiewicz',
+          '你好',
+          '你顥',
           '岩澤',
           '︒',
           'Ｐ',
           '🄟',
           '🐵',
+          '😀',
+          '😁',
         ];
 
         const getSnapshot = await query.get();