@devries-cameron20 / t3code Mirror · read-only
Fork 0

The best way to control Claude Code, Codex and more from any computer or phone

Mirrored from https://github.com/Camo67/t3code.git

Synced Nov 21, 04:45 UTC

main
ts 547 lines 25.2 KB
Raw
1 import JSZip from "jszip";
2 import { openRtf, openNotebook, openDocument, openEpub } from "./extendedDocumentPreview";
3 import type { ScopedThreadRef } from "@t3tools/contracts";
4 import { ProjectReadFileError } from "@t3tools/contracts";
5 import * as Schema from "effect/Schema";
6 import { describe, expect, it, vi } from "vite-plus/test";
7
8 import { promptFileCommentEntries, remapFileCommentAnnotations } from "./fileCommentAnnotations";
9 import {
10 filePreviewReadErrorMessage,
11 isMarkdownPreviewFile,
12 resolveFilePreviewPath,
13 shouldShowFileExplorer,
14 workspaceAssetResource,
15 } from "./filePreviewMode";
16
17 const decodeReadError = Schema.decodeSync(ProjectReadFileError);
18
19 describe("file preview read errors", () => {
20 // regression: #15628 read errors from servers without attempted paths keep their own message
21 it("preserves the public message from older servers", () => {
22 const error = decodeReadError({
23 _tag: "ProjectReadFileError",
24 message: "Legacy file read failure.",
25 });
26 expect(filePreviewReadErrorMessage(error)).toBe("Legacy file read failure.");
27 });
28 });
29
30 describe("file comment annotations", () => {
31 it("keeps an annotation range attached when Pierre remaps its anchor line", () => {
32 expect(
33 remapFileCommentAnnotations([
34 {
35 lineNumber: 20,
36 metadata: {
37 entries: [
38 {
39 id: "comment-1",
40 kind: "comment",
41 startLine: 7,
42 endLine: 16,
43 text: "Keep this guarded.",
44 },
45 ],
46 },
47 },
48 ]),
49 ).toEqual([
50 {
51 lineNumber: 20,
52 metadata: {
53 entries: [
54 {
55 id: "comment-1",
56 kind: "comment",
57 startLine: 11,
58 endLine: 20,
59 text: "Keep this guarded.",
60 },
61 ],
62 },
63 },
64 ]);
65 });
66
67 it("shows only drafts and comments still in the prompt", () => {
68 const entry = (id: string, kind: "draft" | "comment") => ({
69 id,
70 kind,
71 startLine: 3,
72 endLine: 3,
73 text: "",
74 });
75 expect(
76 promptFileCommentEntries(
77 [entry("sent", "comment"), entry("new", "draft"), entry("kept", "comment")],
78 new Set(["kept"]),
79 ),
80 ).toEqual([entry("new", "draft"), entry("kept", "comment")]);
81 });
82 });
83
84 describe("isMarkdownPreviewFile", () => {
85 it("recognizes markdown and MDX files case-insensitively", () => {
86 expect(isMarkdownPreviewFile("README.md")).toBe(true);
87 expect(isMarkdownPreviewFile("docs/guide.MDX")).toBe(true);
88 expect(isMarkdownPreviewFile("docs/guide.MARKDOWN")).toBe(true);
89 expect(isMarkdownPreviewFile("docs/guide.mdown")).toBe(true);
90 expect(isMarkdownPreviewFile("docs/guide.mkd")).toBe(true);
91 });
92 });
93
94 describe("resolveFilePreviewPath", () => {
95 it.each([
96 ["/repo/project", null],
97 ["/repo/project/", null],
98 [".", null],
99 [null, null],
100 ["/repo/project/src", "/repo/project/src"],
101 ["/repo/project/src/main.ts", "/repo/project/src/main.ts"],
102 ["src/main.ts", "src/main.ts"],
103 ["/repo/project-other", "/repo/project-other"],
104 ])("opens %s in the appropriate workspace surface", (path, expected) => {
105 const relativePath = resolveFilePreviewPath(path, "/repo/project");
106 expect(relativePath).toBe(expected);
107 if (expected === null) {
108 expect(
109 shouldShowFileExplorer({ relativePath, explorerOpen: false, attachmentOpen: false }),
110 ).toBe(true);
111 }
112 });
113 });
114
115 describe("workspaceAssetResource", () => {
116 const input = {
117 kind: "workspace-file" as const,
118 threadRef: {
119 environmentId: "env" as ScopedThreadRef["environmentId"],
120 threadId: "thread-1" as ScopedThreadRef["threadId"],
121 },
122 workspaceRoot: "/repo",
123 absolutePath: "/repo/docs/index.html",
124 };
125
126 // regression: #21067 images outside the workspace load through the host media resource
127 it("serves a file outside the workspace as host media and keeps workspace files on workspace access", () => {
128 expect(
129 workspaceAssetResource({ ...input, absolutePath: "/tmp/screenshots/shot.png", draft: false }),
130 ).toEqual({ _tag: "media-file", threadId: "thread-1", path: "/tmp/screenshots/shot.png" });
131 expect(workspaceAssetResource({ ...input, draft: false })).toEqual({
132 _tag: "workspace-file",
133 threadId: "thread-1",
134 path: "/repo/docs/index.html",
135 });
136 });
137
138 it("names the workspace root for a draft, which the server cannot resolve from a thread", () => {
139 expect(workspaceAssetResource({ ...input, draft: true })).toEqual({
140 _tag: "draft-workspace-file",
141 cwd: "/repo",
142 path: "/repo/docs/index.html",
143 });
144 expect(
145 workspaceAssetResource({ ...input, absolutePath: "/tmp/screenshots/shot.png", draft: true }),
146 ).toEqual({ _tag: "draft-workspace-file", cwd: "/repo", path: "/tmp/screenshots/shot.png" });
147 });
148 });
149
150 describe("document preview contents", () => {
151 const bytes = (text: string) => new TextEncoder().encode(text).buffer;
152
153 it.each([
154 ["control symbols", String.raw`{\rtf1 Jean\_Luc\-X\~Y}`, "Jean\u2011Luc\u00adX\u00a0Y"],
155 [
156 "named spaces, joiners and direction marks",
157 String.raw`{\rtf1 A\tab B\emspace C\enspace D\qmspace E\zwj F\zwnj G\zwbo H\ltrmark I\rtlmark J}`,
158 "A\tB\u2003C\u2002D\u2005E\u200dF\u200cG\u200bH\u200eI\u200fJ",
159 ],
160 ["a Unicode fallback", String.raw`{\rtf1 Jean\u8209\_Luc}`, "Jean\u2011Luc"],
161 ])("keeps the text of %s", (_name, source, expected) => {
162 const result = openRtf(bytes(source));
163 if (result.type !== "document") throw new Error("Expected document");
164 expect(result.pages[0]!.html.replaceAll(/<[^>]*>/g, "")).toBe(expected);
165 });
166
167 it.each(["?", String.raw`\'3f`, String.raw`\\`])(
168 "consumes skipped rich text Unicode fallback %s",
169 (fallback) => {
170 const result = openRtf(bytes(String.raw`{\rtf1{\*\unknown\u945` + fallback + "}Hello}"));
171 if (result.type !== "document") throw new Error("Expected document");
172 expect(result.pages[0]!.html).toBe("<p>Hello</p>");
173 },
174 );
175
176 it("bounds EPUB markup made entirely of empty allowed elements", async () => {
177 const archive = new JSZip()
178 .file(
179 "META-INF/container.xml",
180 '<container><rootfiles><rootfile full-path="book.opf"/></rootfiles></container>',
181 )
182 .file(
183 "book.opf",
184 '<package><manifest><item id="a" href="chapter.xhtml" media-type="application/xhtml+xml"/></manifest><spine>' +
185 '<itemref idref="a"/>'.repeat(30) +
186 "</spine></package>",
187 )
188 .file("chapter.xhtml", "<html><body>" + "<figcaption/>".repeat(30_000) + "</body></html>");
189 await expect(openEpub(archive)).rejects.toMatchObject({ code: "rendered-too-large" });
190 });
191
192 it("decodes rich text formatting and Unicode without showing hidden destinations", () => {
193 const result = openRtf(
194 bytes(
195 String.raw`{\rtf1\ansi{\fonttbl{\f0 Hidden font;}}Hello \b bold\b0\par Unicode: \u945?{\*\unknown hidden\_\emspace\zwj}{\field{\*\fldinst HYPERLINK "https://example.com"}{\fldrslt visible}}}`,
196 ),
197 );
198 expect(result.type).toBe("document");
199 if (result.type !== "document") throw new Error("Expected document");
200 const html = result.pages[0]!.html;
201 expect(html.replaceAll(/<[^>]*>/g, "")).toBe("Hello boldUnicode: αvisible");
202 expect(html).toContain("<strong>");
203 expect(html).not.toContain("https:");
204 });
205
206 it("treats a backslash before a line break as a paragraph break, as TextEdit writes them", () => {
207 const html = (rtf: string) => {
208 const result = openRtf(bytes(rtf));
209 if (result.type !== "document") throw new Error("Expected document");
210 return result.pages.map((page) => page.html);
211 };
212 const textEdit = String.raw`{\rtf1\ansi{\fonttbl\f0\fswiss Helvetica;}
213 \pard\f0\fs24 \cf0 First\
214 Second\
215 \
216 Third}`;
217 const expected = html(
218 String.raw`{\rtf1\ansi{\fonttbl\f0\fswiss Helvetica;}\pard\f0\fs24 \cf0 First\par Second\par \par Third}`,
219 );
220 expect(html(textEdit)).toEqual(expected);
221 expect(html(textEdit.replaceAll("\n", "\r\n"))).toEqual(expected);
222 });
223
224 it("preserves rich text page breaks and formatting without splitting hidden destinations", () => {
225 const result = openRtf(
226 bytes(String.raw`{\rtf1\b First\page Second\b0{\*\unknown \page hidden}\page Third\page}`),
227 );
228 if (result.type !== "document") throw new Error("Expected document");
229 expect(result.navigation).toBe("page");
230 expect(result.pages.map((page) => page.html)).toEqual([
231 "<p><strong>First</strong></p>",
232 "<p><strong>Second</strong></p>",
233 "<p>Third</p>",
234 ]);
235 expect(() => openRtf(bytes(String.raw`{\rtf1 ` + String.raw`\page`.repeat(501) + "}"))).toThrow(
236 expect.objectContaining({ code: "rtf-too-many-pages" }),
237 );
238 });
239
240 it("decodes declared rich text code pages for raw bytes and escaped byte runs", () => {
241 const input = new TextEncoder().encode(
242 String.raw`{\rtf1\ansi\ansicpg1251 A \'c0\par\ansicpg932\'82\'a0}`,
243 );
244 input[input.indexOf(65)] = 0xc0;
245 const result = openRtf(input.buffer);
246 if (result.type !== "document") throw new Error("Expected document");
247 expect(result.pages[0]!.html).toBe("<p>А А<br>あ</p>");
248 for (const header of [String.raw`\mac`, String.raw`\mac\ansicpg10000`]) {
249 const mac = openRtf(bytes(String.raw`{\rtf1` + header + String.raw` Caf\'8e Cr\'8fme}`));
250 if (mac.type !== "document") throw new Error("Expected document");
251 expect(mac.pages[0]!.html).toBe("<p>Café Crème</p>");
252 }
253 expect(() => openRtf(bytes(String.raw`{\rtf1\ansicpg99999 text}`))).toThrow(
254 expect.objectContaining({ code: "rtf-code-page" }),
255 );
256 });
257
258 it("resolves scoped ODF namespaces and explicit spreadsheet whitespace", async () => {
259 const archive = new JSZip().file(
260 "content.xml",
261 '<o:document-content xmlns:o="urn:oasis:names:tc:opendocument:xmlns:office:1.0" xmlns:t="urn:oasis:names:tc:opendocument:xmlns:table:1.0"><o:body><o:spreadsheet><t:table t:name="Accounts &amp; names"><t:table-row><t:table-cell><p xmlns="urn:oasis:names:tc:opendocument:xmlns:text:1.0" xmlns:x="urn:oasis:names:tc:opendocument:xmlns:text:1.0">Account<s/>Name<s x:c="3"/>A<tab/>B<line-break/>C &amp; <![CDATA[&amp;]]></p></t:table-cell></t:table-row></t:table></o:spreadsheet></o:body></o:document-content>',
262 );
263 const result = await openDocument(archive, "ods");
264 if (result.type !== "ods") throw new Error("Expected workbook");
265 expect(result.sheets[0]!.name).toBe("Accounts & names");
266 expect(result.sheets[0]!.rows[0]).toEqual(["Account Name A\tB\nC & &amp;"]);
267 });
268
269 it("stops repeated slide images before encoding beyond the output budget", async () => {
270 const pixels = new Uint8Array(4 * 1024 * 1024);
271 pixels.set(
272 Buffer.from(
273 "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVQIHWP4z8DwHwAFgAI/ScLbtAAAAABJRU5ErkJggg==",
274 "base64",
275 ),
276 );
277 const archive = new JSZip()
278 .file(
279 "content.xml",
280 '<o:document-content xmlns:o="urn:oasis:names:tc:opendocument:xmlns:office:1.0" xmlns:d="urn:oasis:names:tc:opendocument:xmlns:drawing:1.0" xmlns:l="http://www.w3.org/1999/xlink"><o:body>' +
281 '<d:page><d:image l:href="picture.png"/></d:page>'.repeat(500) +
282 "</o:body></o:document-content>",
283 )
284 .file("picture.png", pixels);
285 const image = archive.file("picture.png")!;
286 const read = vi.spyOn(image, "async");
287 await expect(openDocument(archive, "odp")).rejects.toMatchObject({
288 code: "rendered-too-large",
289 });
290 expect(read.mock.calls.filter(([type]) => type === "base64")).toHaveLength(2);
291 expect(read).toHaveBeenCalledTimes(5);
292 });
293
294 it("reads saved notebook cells and preferred outputs without executing code", () => {
295 const result = openNotebook(
296 bytes(
297 JSON.stringify({
298 nbformat: 4,
299 metadata: { language_info: { name: "python" } },
300 cells: [
301 { cell_type: "markdown", source: ["# Results"] },
302 {
303 cell_type: "code",
304 source: ["print(42)"],
305 outputs: [
306 { output_type: "stream", text: ["\n "] },
307 { output_type: "display_data", data: {} },
308 { data: { "text/html": [""] } },
309 { data: { "text/html": [], "text/plain": "array fallback" } },
310 { data: { "text/html": "", "text/plain": "\u001b[1mstring fallback\u001b[0m" } },
311 { data: { "text/html": " ", "text/plain": "space fallback" } },
312 { output_type: "stream", text: ["\u001b[", "32m42\u001b[0m\n"] },
313 { data: { "text/html": ["<b>Results</b>"], "text/plain": ["Results"] } },
314 ],
315 },
316 ],
317 }),
318 ),
319 );
320 expect(result).toEqual({
321 type: "notebook",
322 language: "python",
323 cells: [
324 { kind: "markdown", source: "# Results", outputs: [] },
325 {
326 kind: "code",
327 source: "print(42)",
328 outputs: [
329 { kind: "text", text: "array fallback" },
330 { kind: "text", text: "string fallback" },
331 { kind: "text", text: "space fallback" },
332 { kind: "text", text: "42\n" },
333 { kind: "html", text: "<b>Results</b>" },
334 ],
335 },
336 ],
337 });
338 // IPython traceback frames can omit trailing newlines or place ANSI resets after them.
339 const failed = openNotebook(
340 bytes(
341 JSON.stringify({
342 nbformat: 4,
343 cells: [
344 {
345 cell_type: "code",
346 source: "fail()",
347 outputs: [
348 {
349 output_type: "error",
350 traceback: [
351 "-----",
352 "ValueError Traceback (most recent call last)",
353 "Cell In[1], line 4\n----> 4 fail()\n",
354 "Cell In[1], line 2, in fail()\n----> 2 raise ValueError()\n",
355 "ValueError: preview proof",
356 ],
357 },
358 {
359 output_type: "error",
360 traceback: [
361 "Traceback (most recent call last):\n",
362 " Cell In[2], line 5\n fail()\n",
363 "ValueError: preview proof\n",
364 ],
365 },
366 {
367 output_type: "error",
368 traceback: [
369 "\u001b[31mTraceback (most recent call last):\n\u001b[0m",
370 "\u001b[38:2::255:0:0mCell In[3], line 1\n----> 1 fail()\n\u001b[0m\u001b[39m",
371 "\u001b[31mValueError: preview proof\u001b[0m",
372 "final entry",
373 ],
374 },
375 ],
376 },
377 ],
378 }),
379 ),
380 );
381 if (failed.type !== "notebook") throw new Error("Expected notebook");
382 expect(failed.cells[0]!.outputs).toEqual([
383 {
384 kind: "text",
385 text: "-----\nValueError Traceback (most recent call last)\nCell In[1], line 4\n----> 4 fail()\nCell In[1], line 2, in fail()\n----> 2 raise ValueError()\nValueError: preview proof",
386 },
387 {
388 kind: "text",
389 text: "Traceback (most recent call last):\n Cell In[2], line 5\n fail()\nValueError: preview proof\n",
390 },
391 {
392 kind: "text",
393 text: "Traceback (most recent call last):\nCell In[3], line 1\n----> 1 fail()\nValueError: preview proof\nfinal entry",
394 },
395 ]);
396 expect(() => openNotebook(bytes('{"nbformat":3,"cells":[]}'))).toThrow(
397 expect.objectContaining({ code: "notebook-unsupported" }),
398 );
399 expect(() =>
400 openNotebook(
401 bytes(JSON.stringify({ nbformat: 4, cells: Array.from({ length: 501 }, () => ({})) })),
402 ),
403 ).toThrow(expect.objectContaining({ code: "notebook-unsupported" }));
404 });
405
406 it("bounds repeated OpenDocument spreadsheet cells and keeps sheet names and values", async () => {
407 const archive = new JSZip().file(
408 "content.xml",
409 '<office:document-content xmlns:office="urn:oasis:names:tc:opendocument:xmlns:office:1.0" xmlns:table="urn:oasis:names:tc:opendocument:xmlns:table:1.0" xmlns:text="urn:oasis:names:tc:opendocument:xmlns:text:1.0"><office:body><office:spreadsheet><table:table table:name="Budget"><table:table-row table:number-rows-repeated="999999999"><table:table-cell table:number-columns-repeated="999999999"><text:p>Research &amp; design</text:p></table:table-cell></table:table-row></table:table></office:spreadsheet></office:body></office:document-content>',
410 );
411 const result = await openDocument(archive, "ods");
412 expect(result.type).toBe("ods");
413 if (result.type !== "ods") throw new Error("Expected workbook");
414 expect(result.sheets[0]!.name).toBe("Budget");
415 expect(result.sheets[0]!.truncated).toBe(true);
416 expect(result.sheets[0]!.rows).toHaveLength(500);
417 expect(result.sheets[0]!.rows[0]).toHaveLength(50);
418 expect(result.sheets[0]!.rows[0]![0]).toBe("Research & design");
419 });
420
421 it("reads an ODS time cell that has no display paragraph", async () => {
422 const archive = new JSZip().file(
423 "content.xml",
424 `<office:document-content xmlns:office="urn:oasis:names:tc:opendocument:xmlns:office:1.0" xmlns:table="urn:oasis:names:tc:opendocument:xmlns:table:1.0"><office:body><office:spreadsheet><table:table table:name="Values"><table:table-row><table:table-cell office:value-type="time" office:time-value="PT03H30M00S"></table:table-cell></table:table-row></table:table></office:spreadsheet></office:body></office:document-content>`,
425 );
426 const result = await openDocument(archive, "ods");
427 if (result.type !== "ods") throw new Error("Expected workbook");
428 expect(result.sheets[0]!.rows).toEqual([["PT03H30M00S"]]);
429 });
430
431 it.each([1, 500])("ignores styled ODS filler after %s populated rows", async (count) => {
432 const archive = new JSZip().file(
433 "content.xml",
434 `<office:document><table:table><table:table-row table:number-rows-repeated="${count}"><table:table-cell><text:p>North</text:p></table:table-cell><table:table-cell table:number-columns-repeated="1024"/></table:table-row><table:table-row table:number-rows-repeated="1048570"><table:table-cell table:number-columns-repeated="1024"/></table:table-row></table:table></office:document>`,
435 );
436 const result = await openDocument(archive, "ods");
437 if (result.type !== "ods") throw new Error("Expected workbook");
438 expect(result.sheets[0]).toMatchObject({ rowCount: count, columns: 1, truncated: false });
439 expect(result.sheets[0]!.rows).toEqual(Array.from({ length: count }, () => ["North"]));
440 });
441
442 it.each([2, 999999])(
443 "preserves internal ODS gaps of %s rows and reports real overflow",
444 async (gap) => {
445 const archive = new JSZip().file(
446 "content.xml",
447 `<office:document><table:table><table:table-row><table:table-cell><text:p>First</text:p></table:table-cell></table:table-row><table:table-row table:number-rows-repeated="${gap}"><table:table-cell/></table:table-row><table:table-row><table:table-cell table:number-columns-repeated="2"/><table:table-cell><text:p>Last</text:p></table:table-cell></table:table-row></table:table></office:document>`,
448 );
449 const result = await openDocument(archive, "ods");
450 if (result.type !== "ods") throw new Error("Expected workbook");
451 expect(result.sheets[0]).toMatchObject({
452 rowCount: Math.min(gap + 2, 500),
453 columns: gap === 2 ? 3 : 1,
454 truncated: gap !== 2,
455 });
456 expect(result.sheets[0]!.rows[1]).toEqual([]);
457 if (gap === 2) expect(result.sheets[0]!.rows[3]).toEqual(["", "", "Last"]);
458 },
459 );
460
461 const approved = "<text:p>Approved <text:span>now</text:span></text:p>";
462 const approvedRow = `<tr>${"<td><p>Approved <span>now</span></p></td>".repeat(2)}</tr>`;
463 it.each([
464 [
465 "expands",
466 `<table:table-row>${`<table:table-cell>${approved}</table:table-cell>`.repeat(2)}</table:table-row><table:table-row table:number-rows-repeated="2"><table:table-cell table:number-columns-repeated="2">${approved}</table:table-cell></table:table-row>`,
467 approvedRow.repeat(3),
468 ],
469 [
470 "bounds",
471 '<table:table-row table:number-rows-repeated="999999999"><table:table-cell table:number-columns-repeated="999999999"><text:p>x</text:p></table:table-cell></table:table-row>',
472 `<tr>${"<td><p>x</p></td>".repeat(50)}</tr>`.repeat(500),
473 ],
474 ])("%s repeated ODT table rows and cells", async (_, rows, html) => {
475 const archive = new JSZip().file(
476 "content.xml",
477 `<office:document-content xmlns:office="urn:oasis:names:tc:opendocument:xmlns:office:1.0" xmlns:table="urn:oasis:names:tc:opendocument:xmlns:table:1.0" xmlns:text="urn:oasis:names:tc:opendocument:xmlns:text:1.0"><office:body><office:text><table:table><table:table-column table:number-columns-repeated="2"/>${rows}</table:table></office:text></office:body></office:document-content>`,
478 );
479 const result = await openDocument(archive, "odt");
480 if (result.type !== "document") throw new Error("Expected document");
481 expect(result.pages[0]!.html).toBe(`<table>${html}</table>`);
482 });
483
484 it.each(["odt", "odp"] as const)("reads %s text and rejects unsupported XML", async (kind) => {
485 const archive = new JSZip().file(
486 "content.xml",
487 `<office:document-content xmlns:office="urn:oasis:names:tc:opendocument:xmlns:office:1.0" xmlns:text="urn:oasis:names:tc:opendocument:xmlns:text:1.0" xmlns:draw="urn:oasis:names:tc:opendocument:xmlns:drawing:1.0" xmlns:presentation="urn:oasis:names:tc:opendocument:xmlns:presentation:1.0"><office:body>${kind === "odt" ? "<office:text><text:p>Field notes</text:p></office:text>" : '<draw:page draw:name="Overview"><text:h>Launch plan</text:h><presentation:notes><draw:frame><draw:text-box><text:p>Speaker only</text:p></draw:text-box></draw:frame></presentation:notes></draw:page>'}</office:body></office:document-content>`,
488 );
489 const result = await openDocument(archive, kind);
490 expect(result.type).toBe("document");
491 if (result.type !== "document") throw new Error("Expected document");
492 expect(result.pages[0]!.html).toContain(kind === "odt" ? "Field notes" : "Launch plan");
493 expect(result.pages[0]!.html).not.toContain("Speaker only");
494 for (const source of [
495 '<!DOCTYPE x [<!ENTITY external SYSTEM "file:///etc/passwd">]><x>&external;</x>',
496 '<html><body><p title="<!--">x</p><!DOCTYPE x [<!ENTITY e "boom">]><b title="-->">&e;</b></body></html>',
497 '<html><body><p title="<![CDATA[">x</p><!DOCTYPE x [<!ENTITY e "boom">]><b title="]]>">&e;</b></body></html>',
498 '<!DOC<!DOCTYPE a>TYPE x [<!EN<!DOCTYPE a>TITY e "boom">]><x>&e;</x>',
499 '<?test <![CDATA[ ?><!DOCTYPE html [<!ENTITY e "boom">]><html><body><pre><![CDATA[x]]></pre></body></html>',
500 '<?test <!-- ?><!DOCTYPE html [<!ENTITY e "boom">]><html><body><!-- x --></body></html>',
501 "<html><body>" + "<!--".repeat(80_000) + "</body></html>",
502 "<html><body>" + "<![CDATA[".repeat(80_000) + "</body></html>",
503 ]) {
504 archive.file("content.xml", source);
505 await expect(openDocument(archive, kind)).rejects.toMatchObject({ code: "xml-invalid" });
506 }
507 });
508
509 it("reads EPUB chapters in spine order and removes active and remote content", async () => {
510 const archive = new JSZip()
511 .file(
512 "META-INF/container.xml",
513 '<container><rootfiles><rootfile full-path="Book/book.opf"/></rootfiles></container>',
514 )
515 .file(
516 "Book/book.opf",
517 '<package><manifest><item id="a" href="one.xhtml" media-type="application/xhtml+xml"/><item id="b" href="two&amp;&#x26;&#38;.xhtml" media-type="application/xhtml+xml"/></manifest><spine><itemref idref="b"/><itemref idref="a"/></spine></package>',
518 )
519 .file(
520 "Book/one.xhtml",
521 '<html><head><title>First</title></head><body><h1>First chapter</h1><img src="cover.png" alt="Cover"/><img src="https://example.com/tracker.png"/><script>alert(1)</script><style>@import "https://example.com/style.css";</style></body></html>',
522 )
523 .file(
524 "Book/two&&&.xhtml",
525 '<?xml-stylesheet href="a<b.css"?><!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd"><html><head><title>Second</title></head><body><h1>Second chapter</h1><p>A&nbsp;B &copy; &mdash; &NotEqualTilde; &amp;lt;</p><pre><![CDATA[Use &amp; and &lt; literally]]></pre><pre><![CDATA[<!DOCTYPE html><!ENTITY author "Ada">]]></pre><!-- <!ENTITY note "x"> --><img src="cover.png" alt="&lt; &copy; &NotEqualTilde;"/></body></html>',
526 )
527 .file(
528 "Book/cover.png",
529 Buffer.from(
530 "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVQIHWP4z8DwHwAFgAI/ScLbtAAAAABJRU5ErkJggg==",
531 "base64",
532 ),
533 );
534 const result = await openEpub(archive);
535 expect(result.type).toBe("document");
536 if (result.type !== "document") throw new Error("Expected document");
537 expect(result.pages.map((page) => page.title)).toEqual(["Second", "First"]);
538 expect(result.pages[0]!.html).toContain("A\u00a0B © — ≂̸ &amp;lt;");
539 expect(result.pages[0]!.html).toContain("<pre>Use &amp;amp; and &amp;lt; literally</pre>");
540 expect(result.pages[0]!.html).toContain(
541 "<pre>&lt;!DOCTYPE html&gt;&lt;!ENTITY author &quot;Ada&quot;&gt;</pre>",
542 );
543 expect(result.pages[0]!.html).toContain('alt="&lt; © ≂̸"');
544 expect(result.pages[1]!.html).toContain('src="data:image/png;base64,');
545 expect(result.pages[1]!.html).not.toMatch(/https:|script|alert/);
546 });
547 });