Files
readest/apps/readest-app/src/__tests__/utils/xcfi.spec.ts
T
Huang Xin 3a81e09911 fix(reader): scroll oversized blocks in-place instead of turning the page (#4400) (#4415)
Wide or tall tables, code blocks and display equations overflowed the reading
column and a scroll gesture over them turned the page instead of scrolling the
content (#4400).

- Wrap tables and display equations in a horizontally/vertically scrollable
  container; route touch + wheel along the box's scrollable axis so it scrolls
  the box and never turns the page, even at the edge (both axes).
- A box that fits its column is marked fit (overflow:visible) so it never clips
  or captures gestures; the fit decision is measured once after layout via a
  self-disconnecting ResizeObserver, so it never relayerizes during a page turn.
- The scroll wrapper carries a new cfi-skip attribute that makes it transparent
  to CFI: epubcfi.js hoists a cfi-skip node's children into its parent (unlike
  cfi-inert which drops the subtree), and xcfi.ts mirrors this for CFI<->XPointer
  so existing highlights, bookmarks and KOSync positions inside a wrapped table
  or equation still resolve. The sanitizer whitelists cfi-skip.
- Bump foliate-js submodule (cfi-skip support + raf fallback for large sections).

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-02 14:45:18 +02:00

600 lines
22 KiB
TypeScript

import { describe, it, expect, beforeEach } from 'vitest';
import { XCFI } from '@/utils/xcfi';
describe('CFIToXPointerConverter', () => {
let converter: XCFI;
let simpleDoc: Document;
let complexDoc: Document;
beforeEach(() => {
simpleDoc = new DOMParser().parseFromString(
`
<html>
<head>
<title>Simple Document</title>
</head>
<body>
<div>
<p>First paragraph</p>
<p>Second paragraph with some text</p>
<p>Third paragraph</p>
</div>
</body>
</html>
`,
'text/html',
);
complexDoc = new DOMParser().parseFromString(
`
<html>
<head>
<title>Complex Document</title>
</head>
<body>
<section>
<h1>Chapter 1</h1>
<p>First paragraph</p>
<p>Second paragraph</p>
</section>
<section>
<h1>Chapter 2</h1>
<p id="special">Another paragraph</p>
<p>Final paragraph with <em>emphasis</em> and more text</p>
</section>
</body>
</html>
`,
'text/html',
);
});
describe('static methods', () => {
it('should extract spine index from CFI', () => {
const cfi1 = 'epubcfi(/6/2!/4/2/4)'; // Spine index 0
const cfi2 = 'epubcfi(/6/4!/4/2/4)'; // Spine index 1
const cfi3 = 'epubcfi(/6/10!/4/2/4)'; // Spine index 4
expect(XCFI.extractSpineIndex(cfi1)).toBe(0);
expect(XCFI.extractSpineIndex(cfi2)).toBe(1);
expect(XCFI.extractSpineIndex(cfi3)).toBe(4);
});
it('should extract spine index from range CFI', () => {
const rangeCfi = 'epubcfi(/6/8!/4/2/2/1:5,/6/8!/4/2/4/1:10)'; // Spine index 3
expect(XCFI.extractSpineIndex(rangeCfi)).toBe(3);
});
it('should extract spine index from CFI with assertions', () => {
const cfi = 'epubcfi(/6/1266!/4,/76,/88/1:85)'; // Complex CFI, spine index 632
expect(XCFI.extractSpineIndex(cfi)).toBe(632);
});
it('should throw error for invalid CFI in extractSpineIndex', () => {
const invalidCfi = 'invalid-cfi';
expect(() => XCFI.extractSpineIndex(invalidCfi)).toThrow('Cannot extract spine index');
});
});
describe('round-trip conversion - point CFI', () => {
beforeEach(() => {
converter = new XCFI(simpleDoc, 1);
});
it('should convert first element CFI round-trip', () => {
const originalCfi = 'epubcfi(/6/4!/4/2/2)'; // First p element
const xpointer = converter.cfiToXPointer(originalCfi);
const convertedCfi = converter.xPointerToCFI(xpointer.xpointer);
expect(originalCfi).toEqual(convertedCfi);
expect(xpointer).toEqual({
xpointer: '/body/DocFragment[2]/body/div/p[1]',
});
});
it('should convert basic element CFI round-trip', () => {
const originalCfi = 'epubcfi(/6/4!/4/2/4)'; // Second p element
const xpointer = converter.cfiToXPointer(originalCfi);
const convertedCfi = converter.xPointerToCFI(xpointer.xpointer);
expect(originalCfi).toEqual(convertedCfi);
expect(xpointer).toEqual({
xpointer: '/body/DocFragment[2]/body/div/p[2]',
});
});
it('should convert third element CFI round-trip', () => {
const originalCfi = 'epubcfi(/6/4!/4/2/6)'; // Third p element
const xpointer = converter.cfiToXPointer(originalCfi);
const convertedCfi = converter.xPointerToCFI(xpointer.xpointer);
expect(originalCfi).toEqual(convertedCfi);
expect(xpointer).toEqual({
xpointer: '/body/DocFragment[2]/body/div/p[3]',
});
});
});
describe('round-trip conversion - range CFI', () => {
beforeEach(() => {
converter = new XCFI(simpleDoc, 2);
});
it('should convert standard range CFI', () => {
const originalCfi = 'epubcfi(/6/6!/4/2,/2/1:6,/4/1:16)'; // From first p:6 to second p:16
const xpointer = converter.cfiToXPointer(originalCfi);
const convertedCfi = converter.xPointerToCFI(xpointer.pos0!, xpointer.pos1!);
expect(originalCfi).toEqual(convertedCfi);
expect(xpointer.xpointer).toEqual('/body/DocFragment[3]/body/div/p[1]/text().6');
expect(xpointer.pos0).toEqual('/body/DocFragment[3]/body/div/p[1]/text().6');
expect(xpointer.pos1).toEqual('/body/DocFragment[3]/body/div/p[2]/text().16');
});
it('should convert range CFI within same element', () => {
const originalCfi = 'epubcfi(/6/6!/4/2/4,/1:5,/1:10)'; // Within second p element
const xpointer = converter.cfiToXPointer(originalCfi);
const convertedCfi = converter.xPointerToCFI(xpointer.pos0!, xpointer.pos1!);
expect(originalCfi).toEqual(convertedCfi);
expect(xpointer.pos0).toMatch(/\/text\(\)\.5$/);
expect(xpointer.pos1).toMatch(/\/text\(\)\.10$/);
});
it('should handle range across multiple elements', () => {
const originalCfi = 'epubcfi(/6/6!/4/2,/2,/6)'; // From first to third p
const xpointer = converter.cfiToXPointer(originalCfi);
const convertedCfi = converter.xPointerToCFI(xpointer.pos0!, xpointer.pos1!);
expect(originalCfi).toEqual(convertedCfi);
expect(xpointer.pos0).toMatch(/p\[1\]/);
expect(xpointer.pos1).toMatch(/p\[3\]/);
});
});
describe('round-trip conversion - complex document', () => {
beforeEach(() => {
converter = new XCFI(complexDoc, 3);
});
it('should handle nested elements', () => {
const originalCfi = 'epubcfi(/6/8!/4/2/2)'; // First section
const xpointer = converter.cfiToXPointer(originalCfi);
const convertedCfi = converter.xPointerToCFI(xpointer.xpointer);
expect(originalCfi).toEqual(convertedCfi);
expect(xpointer.xpointer).toMatch(/\/body\/section\[1\]/);
});
it('should handle elements with IDs', () => {
const originalCfi = 'epubcfi(/6/8!/4/4/4[special])'; // Element with id="special"
const xpointer = converter.cfiToXPointer(originalCfi);
const convertedCfi = converter.xPointerToCFI(xpointer.xpointer);
expect(originalCfi).toEqual(convertedCfi);
expect(xpointer.xpointer).toMatch(/\/body\/section\[2\]\/p\[1\]/);
});
it('should handle inline elements', () => {
const originalCfi = 'epubcfi(/6/8!/4/4/6)'; // Text with inline em element
const xpointer = converter.cfiToXPointer(originalCfi);
const convertedCfi = converter.xPointerToCFI(xpointer.xpointer);
expect(originalCfi).toEqual(convertedCfi);
expect(xpointer.xpointer).toMatch(/\/body\/section\[2\]\/p\[2\]/);
});
});
describe('convertCFI - error handling', () => {
beforeEach(() => {
converter = new XCFI(simpleDoc, 0);
});
it('should throw error for invalid CFI format', () => {
const invalidCfi = 'invalid-cfi';
expect(() => converter.cfiToXPointer(invalidCfi)).toThrow('Failed to convert CFI');
});
it('should throw error for CFI with invalid path', () => {
const invalidCfi = 'epubcfi(/6/999!/2/2)'; // Non-existent path
expect(() => converter.cfiToXPointer(invalidCfi)).toThrow();
});
it('should handle malformed CFI gracefully', () => {
const malformedCfi = 'epubcfi(/6/2/2';
expect(() => converter.cfiToXPointer(malformedCfi)).toThrow();
});
});
describe('xPointerToCFI - direct XPointer input', () => {
beforeEach(() => {
converter = new XCFI(simpleDoc, 1);
});
it('should convert XPointer to CFI for first element', () => {
const xpointer = '/body/DocFragment[2]/body/div/p[1]';
const cfi = converter.xPointerToCFI(xpointer);
// Verify by converting back to XPointer
const backToXPointer = converter.cfiToXPointer(cfi);
expect(backToXPointer.xpointer).toBe(xpointer);
});
it('should convert XPointer to CFI for second element', () => {
const xpointer = '/body/DocFragment[2]/body/div/p[2]';
const cfi = converter.xPointerToCFI(xpointer);
const backToXPointer = converter.cfiToXPointer(cfi);
expect(backToXPointer.xpointer).toBe(xpointer);
});
it('should convert XPointer with text offset to CFI', () => {
const xpointer = '/body/DocFragment[2]/body/div[0]/p[1]/text().6';
const cfi = converter.xPointerToCFI(xpointer);
expect(cfi).toBe('epubcfi(/6/4!/4/2/2/1:6)');
});
it('should convert range XPointer to CFI', () => {
const pos0 = '/body/DocFragment[2]/body/div/p[1]/text().6';
const pos1 = '/body/DocFragment[2]/body/div/p[2]/text().16';
const cfi = converter.xPointerToCFI(pos0, pos1);
const xpointer = converter.cfiToXPointer(cfi);
expect(cfi).toMatch(/^epubcfi\([^,]+,[^,]+,[^,]+\)$/);
// cfiToXPointer now produces text()[K].N format
expect(xpointer.pos0).toBe('/body/DocFragment[2]/body/div/p[1]/text().6');
expect(xpointer.pos1).toBe('/body/DocFragment[2]/body/div/p[2]/text().16');
// Round-trip: the new format should convert back to the same CFI
const backCfi = converter.xPointerToCFI(xpointer.pos0!, xpointer.pos1!);
expect(backCfi).toBe(cfi);
});
});
describe('xPointerToCFI - error handling', () => {
beforeEach(() => {
converter = new XCFI(simpleDoc, 0);
});
it('should throw error for invalid XPointer format', () => {
const invalidXPointer = 'invalid-xpointer';
expect(() => converter.xPointerToCFI(invalidXPointer)).toThrow('Failed to convert XPointer');
});
it('should throw error for XPointer with non-existent path', () => {
const invalidXPointer = '/body/DocFragment[1]/body/nonexistent[999]';
expect(() => converter.xPointerToCFI(invalidXPointer)).toThrow();
});
it('should throw error for malformed XPointer', () => {
const malformedXPointer = '/body/DocFragment[1]/body/div[';
expect(() => converter.xPointerToCFI(malformedXPointer)).toThrow();
});
it('should handle CFI without spine step prefix', () => {
// Test the adjustSpineIndex method handles CFIs that don't start with /6/n!
const converter = new XCFI(simpleDoc, 3); // Use different spine index
const xpointer = '/body/DocFragment[4]/body/div/p[1]';
const cfi = converter.xPointerToCFI(xpointer);
// Verify the spine step is correctly added/adjusted
expect(cfi).toMatch(/^epubcfi\(\/6\/8!/); // (3+1)*2 = 8
// Verify round-trip works
const backToXPointer = converter.cfiToXPointer(cfi);
expect(backToXPointer.xpointer).toBe(xpointer);
});
});
describe('validateCFI', () => {
beforeEach(() => {
converter = new XCFI(simpleDoc, 0);
});
it('should validate correct CFI', () => {
const validCfi = 'epubcfi(/6/2!/4/4)';
expect(converter.validateCFI(validCfi)).toBe(true);
});
it('should invalidate incorrect CFI format', () => {
const invalidCfi = 'invalid-cfi';
expect(converter.validateCFI(invalidCfi)).toBe(false);
});
it('should invalidate CFI with wrong path', () => {
const invalidCfi = 'epubcfi(/6/2!/998/2/2)';
expect(converter.validateCFI(invalidCfi)).toBe(false);
});
it('should validate range CFI', () => {
const validRangeCfi = 'epubcfi(/6/2!/4/2,/2/1:5,/4/1:10)';
expect(converter.validateCFI(validRangeCfi)).toBe(true);
});
});
describe('edge cases', () => {
beforeEach(() => {
converter = new XCFI(simpleDoc, 0);
});
it('should handle empty elements', () => {
const emptyDoc = new DOMParser().parseFromString(
`
<html>
<body>
<div>
<p></p>
<p>Non-empty</p>
</div>
</body>
</html>
`,
'text/html',
);
const converter = new XCFI(emptyDoc, 2);
const cfi = 'epubcfi(/6/6!/4/2/2)'; // Empty p element
const result = converter.cfiToXPointer(cfi);
expect(result.xpointer).toBe('/body/DocFragment[3]/body/div/p[1]');
});
it('should handle whitespace-only text nodes', () => {
const whitespaceDoc = new DOMParser().parseFromString(
`
<html>
<body>
<div>
<p> </p>
<p>Real content</p>
</div>
</body>
</html>
`,
'text/html',
);
const converter = new XCFI(whitespaceDoc, 2);
const cfi = 'epubcfi(/6/6!/4/2/4)'; // Second p element
const result = converter.cfiToXPointer(cfi);
expect(result.xpointer).toBe('/body/DocFragment[3]/body/div/p[2]');
});
it('should handle deeply nested elements', () => {
const nestedDoc = new DOMParser().parseFromString(
`
<html>
<body>
<div>
<section>
<article>
<p>Deeply nested p0</p>
<p>Deeply nested p1</p>
</article>
</section>
</div>
</body>
</html>
`,
'text/html',
);
const converter = new XCFI(nestedDoc, 2);
const cfi = 'epubcfi(/6/6!/4/2/2/2/2)'; // Deeply nested p
const result = converter.cfiToXPointer(cfi);
expect(result.xpointer).toBe('/body/DocFragment[3]/body/div/section/article/p[1]');
});
});
describe('indexed text node XPointers (text()[N].offset)', () => {
let inlineDoc: Document;
beforeEach(() => {
// Simulates: <p>...text...<a id="page96"></a> Megan likes the sea, too...</p>
inlineDoc = new DOMParser().parseFromString(
`<html><body>
<p>She spent a year of her Ph.D. at my old college at Cambridge. <a id="page96" tabindex="-1"></a>A woman, at Caius! Megan likes the sea, too. She's finishing her radioastronomy research.</p>
</body></html>`,
'text/html',
);
});
it('should convert text()[N].offset range XPointer to valid CFI', () => {
const converter = new XCFI(inlineDoc, 10);
// text()[2] = the 2nd direct text node child of <p>, i.e. the text after the <a>
const pos0 = '/body/DocFragment[11]/body/p/text()[2].44';
const pos1 = '/body/DocFragment[11]/body/p/text()[2].69';
const cfi = converter.xPointerToCFI(pos0, pos1);
// Should produce a valid range CFI pointing into the 3rd child node (text after <a>)
expect(cfi).toMatch(/^epubcfi\(/);
expect(cfi).toMatch(/,.*,/); // Range CFI has two commas
// /3 = 3rd child of <p> (1:text, 2:<a>, 3:text), offsets 44 and 69
expect(cfi).toContain('/3:44');
expect(cfi).toContain('/3:69');
});
it('should convert text()[1].offset XPointer to valid CFI', () => {
const converter = new XCFI(inlineDoc, 10);
// text()[1] = the 1st direct text node child of <p>, i.e. text before the <a>
const xp = '/body/DocFragment[11]/body/p/text()[1].5';
const cfi = converter.xPointerToCFI(xp);
expect(cfi).toMatch(/^epubcfi\(/);
// /1 = 1st child of <p> (text node), offset 5
expect(cfi).toContain('/1:5');
});
it('should handle text()[N].offset with multiple inline elements', () => {
const multiInlineDoc = new DOMParser().parseFromString(
`<html><body>
<p>Start text <em>emphasis</em> middle text <a id="link1">link</a> end text here.</p>
</body></html>`,
'text/html',
);
const converter = new XCFI(multiInlineDoc, 5);
// Direct children of <p>: text, <em>, text, <a>, text
// text()[3] = " end text here." (3rd direct text node of <p>)
const xp = '/body/DocFragment[6]/body/p/text()[3].4';
const cfi = converter.xPointerToCFI(xp);
expect(cfi).toMatch(/^epubcfi\(/);
// /5 = 5th child of <p> (1:text, 2:<em>, 3:text, 4:<a>, 5:text), offset 4
expect(cfi).toContain('/5:4');
});
it('should produce correct CFI for text()[1] with no inline siblings', () => {
const simpleDoc = new DOMParser().parseFromString(
`<html><body><p>Hello world</p></body></html>`,
'text/html',
);
const converter = new XCFI(simpleDoc, 0);
const xp = '/body/DocFragment[1]/body/p/text()[1].5';
const cfi = converter.xPointerToCFI(xp);
expect(cfi).toMatch(/^epubcfi\(/);
expect(cfi).toContain('/1:5');
});
});
describe('cfi-inert elements should be invisible to XPointer', () => {
it('should skip cfi-inert div when resolving KOReader XPointer', () => {
const doc = new DOMParser().parseFromString(
`<html><body>
<div cfi-inert="">skip link</div>
<div class="body">
<div class="chapter">
<div class="text">
<p>Alice thought this a very curious thing.</p>
</div>
</div>
</div>
</body></html>`,
'text/html',
);
const converter = new XCFI(doc, 10);
// KOReader XPointer: div (no index) means the only "real" div
const xp = '/body/DocFragment[11]/body/div/div/div/p/text().10';
const cfi = converter.xPointerToCFI(xp);
expect(cfi).toMatch(/^epubcfi\(/);
});
it('should skip cfi-inert div when building XPointer path from element', () => {
const doc = new DOMParser().parseFromString(
`<html><body>
<div cfi-inert="">skip link</div>
<div class="body">
<div class="chapter">
<div class="text">
<p>Alice thought this a very curious thing.</p>
</div>
</div>
</div>
</body></html>`,
'text/html',
);
const converter = new XCFI(doc, 10);
// Navigate to the <p> via DOM, then build XPointer from it
const p = doc.querySelector('p')!;
// Use xPointerToCFI and verify the KOReader XPointer resolves correctly
const koXp = '/body/DocFragment[11]/body/div/div/div/p/text().10';
const cfi = converter.xPointerToCFI(koXp);
// Verify the same CFI is produced when starting from a Readest-generated range
const range = doc.createRange();
const textNode = p.firstChild!;
range.setStart(textNode, 10);
range.setEnd(textNode, 10);
// The xPointerToCFI should resolve through the content div, not the cfi-inert div
expect(cfi).toMatch(/^epubcfi\(/);
});
it('should handle cfi-inert with multiple real siblings', () => {
const doc = new DOMParser().parseFromString(
`<html><body>
<div cfi-inert="">skip</div>
<div class="first">First</div>
<div class="second"><p>Content</p></div>
</body></html>`,
'text/html',
);
const converter = new XCFI(doc, 0);
// Two real divs: div[1]=first, div[2]=second — cfi-inert is invisible
const xp = '/body/DocFragment[1]/body/div[2]/p';
const cfi = converter.xPointerToCFI(xp);
expect(cfi).toMatch(/^epubcfi\(/);
});
});
// applyScrollableStyle wraps wide tables/equations in a cfi-skip <div>. KOReader's
// CREngine DOM has no such wrapper, so CFI ↔ XPointer must treat the wrapper as
// transparent: the same KOReader XPointer must map to the same CFI (and back) with
// or without the wrapper — otherwise KOSync positions would drift.
describe('cfi-skip wrapper is invisible to CFI <-> XPointer', () => {
// identical content; only difference is the wrapping cfi-skip div around the table
const inner = `<table><tbody><tr><td><p>Hello scrollable world</p></td></tr></tbody></table>`;
const html = (table: string) =>
`<html><body><p>Intro paragraph</p>${table}<p>Outro paragraph</p></body></html>`;
const baseDoc = () => new DOMParser().parseFromString(html(inner), 'text/html');
const wrappedDoc = () =>
new DOMParser().parseFromString(
html(`<div class="scroll-wrapper" cfi-skip="">${inner}</div>`),
'text/html',
);
// KOReader-style XPointers (no wrapper) — the wrapper-less DOM CREngine produces.
const cellText = '/body/DocFragment[1]/body/table/tbody/tr/td/p/text().5';
const tableElement = '/body/DocFragment[1]/body/table';
const afterTable = '/body/DocFragment[1]/body/p[2]/text().3';
it('maps a KOReader XPointer to the same CFI with and without the wrapper', () => {
const base = new XCFI(baseDoc(), 0);
const wrapped = new XCFI(wrappedDoc(), 0);
for (const xp of [cellText, tableElement, afterTable]) {
// The wrapper must not change the CFI the XPointer resolves to.
expect(wrapped.xPointerToCFI(xp), xp).toBe(base.xPointerToCFI(xp));
}
});
it('round-trips the table element XPointer back to itself through the wrapped doc', () => {
const wrapped = new XCFI(wrappedDoc(), 0);
const cfi = wrapped.xPointerToCFI(tableElement);
const built = wrapped.cfiToXPointer(cfi);
expect(built.xpointer).toBe(tableElement);
// The wrapper is a <div>; a transparent wrapper must not appear in the path.
expect(built.xpointer).not.toContain('/div');
});
it('round-trips a KOReader range (highlight) inside the table unchanged', () => {
// A highlight is a range; xPointerToCFI(pos0,pos1) -> cfiToXPointer pos0/pos1
// exercises rangePointToXPointer, the path real highlights use.
const pos0 = '/body/DocFragment[1]/body/table/tbody/tr/td/p/text().0';
const pos1 = '/body/DocFragment[1]/body/table/tbody/tr/td/p/text().5';
const base = new XCFI(baseDoc(), 0);
const wrapped = new XCFI(wrappedDoc(), 0);
const baseCfi = base.xPointerToCFI(pos0, pos1);
const wrappedCfi = wrapped.xPointerToCFI(pos0, pos1);
expect(wrappedCfi).toBe(baseCfi); // wrapper doesn't change the highlight CFI
const back = wrapped.cfiToXPointer(wrappedCfi);
expect(back.pos0).toBe(pos0);
expect(back.pos1).toBe(pos1);
expect(back.pos0).not.toContain('/div');
expect(back.pos1).not.toContain('/div');
});
it('builds the same CFI->XPointer for the table element with and without the wrapper', () => {
const base = new XCFI(baseDoc(), 0);
const wrapped = new XCFI(wrappedDoc(), 0);
const cfi = base.xPointerToCFI(tableElement);
expect(wrapped.cfiToXPointer(cfi).xpointer).toBe(base.cfiToXPointer(cfi).xpointer);
});
});
});