refactor: store invisible characters as escapes in registry and tests

INVISIBLE_CHARACTERS held the six characters as raw literals while a
neighboring comment claimed the module used escapes to keep the source
readable. Switch the registry and its test file to explicit \uXXXX escapes
so the comment matches reality and each code point is verifiable in a diff.

Author: dkrizan

webapp/src/fixtures/__tests__/invisibleCharacters.test.ts

@@ -30,52 +30,52 @@ describe('findInvisibleCharacters', () => {
   });
 
   it.each([
-    [' ', 'nonBreakingSpace'],
-    [' ', 'nonBreakingSpace'],
-    [' ', 'nonBreakingSpace'],
-    ['​', 'zeroWidth'],
-    ['', 'zeroWidth'],
-    ['­', 'zeroWidth'],
+    ['\u00A0', 'nonBreakingSpace'],
+    ['\u202F', 'nonBreakingSpace'],
+    ['\u2007', 'nonBreakingSpace'],
+    ['\u200B', 'zeroWidth'],
+    ['\uFEFF', 'zeroWidth'],
+    ['\u00AD', 'zeroWidth'],
   ])('finds %j and reports kind %s', (value, kind) => {
     expect(findInvisibleCharacters(`a${value}b`)).toEqual([
       { index: 1, char: { value, kind } },
     ]);
   });
 
   it('finds a character at the first index', () => {
-    expect(findInvisibleCharacters(' abc')).toEqual([
-      { index: 0, char: { value: ' ', kind: 'nonBreakingSpace' } },
+    expect(findInvisibleCharacters('\u00A0abc')).toEqual([
+      { index: 0, char: { value: '\u00A0', kind: 'nonBreakingSpace' } },
     ]);
   });
 
   it('finds a character at the last index', () => {
-    expect(findInvisibleCharacters('abc ')).toEqual([
-      { index: 3, char: { value: ' ', kind: 'nonBreakingSpace' } },
+    expect(findInvisibleCharacters('abc\u00A0')).toEqual([
+      { index: 3, char: { value: '\u00A0', kind: 'nonBreakingSpace' } },
     ]);
   });
 
   it('finds multiple characters in ascending order', () => {
-    expect(findInvisibleCharacters('a b​c')).toEqual([
-      { index: 1, char: { value: ' ', kind: 'nonBreakingSpace' } },
-      { index: 3, char: { value: '​', kind: 'zeroWidth' } },
+    expect(findInvisibleCharacters('a\u00A0b\u200Bc')).toEqual([
+      { index: 1, char: { value: '\u00A0', kind: 'nonBreakingSpace' } },
+      { index: 3, char: { value: '\u200B', kind: 'zeroWidth' } },
     ]);
   });
 
   it('finds adjacent characters', () => {
-    expect(findInvisibleCharacters('  ')).toEqual([
-      { index: 0, char: { value: ' ', kind: 'nonBreakingSpace' } },
-      { index: 1, char: { value: ' ', kind: 'nonBreakingSpace' } },
+    expect(findInvisibleCharacters('\u00A0\u00A0')).toEqual([
+      { index: 0, char: { value: '\u00A0', kind: 'nonBreakingSpace' } },
+      { index: 1, char: { value: '\u00A0', kind: 'nonBreakingSpace' } },
     ]);
   });
 
   it('reports code-unit offsets in text containing surrogate pairs', () => {
-    expect(findInvisibleCharacters('\u{1f600} ')).toEqual([
-      { index: 2, char: { value: ' ', kind: 'nonBreakingSpace' } },
+    expect(findInvisibleCharacters('\u{1f600}\u00A0')).toEqual([
+      { index: 2, char: { value: '\u00A0', kind: 'nonBreakingSpace' } },
     ]);
   });
 
   it('is stable across repeated calls', () => {
-    const text = 'a b';
+    const text = 'a\u00A0b';
     expect(findInvisibleCharacters(text)).toEqual(
       findInvisibleCharacters(text)
     );

webapp/src/fixtures/invisibleCharacters.ts

@@ -6,12 +6,12 @@ export type InvisibleChar = {
 };
 
 export const INVISIBLE_CHARACTERS: InvisibleChar[] = [
-  { value: ' ', kind: 'nonBreakingSpace' },
-  { value: ' ', kind: 'nonBreakingSpace' },
-  { value: ' ', kind: 'nonBreakingSpace' },
-  { value: '​', kind: 'zeroWidth' },
-  { value: '', kind: 'zeroWidth' },
-  { value: '­', kind: 'zeroWidth' },
+  { value: '\u00A0', kind: 'nonBreakingSpace' },
+  { value: '\u202F', kind: 'nonBreakingSpace' },
+  { value: '\u2007', kind: 'nonBreakingSpace' },
+  { value: '\u200B', kind: 'zeroWidth' },
+  { value: '\uFEFF', kind: 'zeroWidth' },
+  { value: '\u00AD', kind: 'zeroWidth' },
 ];
 
 const BY_VALUE = new Map(