pdf.js/test/unit/writer_spec.js
Calixte Denizet bc87f4e8d6 Add the possibility to create a pdf from different ones (bug 1997379)
For now it's just possible to create a single pdf in selecting some pages in different pdf sources.
The merge is for now pretty basic (it's why it's still a WIP) none of these data are merged for now:
 - the struct trees
 - the page labels
 - the outlines
 - named destinations
For there are 2 new ref tests where some new pdfs are created: one with some extracted pages and an other
one (encrypted) which is just rewritten.
The ref images are generated from the original pdfs in selecting the page we want and the new images are
taken from the generated pdfs.
2025-11-07 14:57:48 +01:00

330 lines
9.0 KiB
JavaScript

/* Copyright 2020 Mozilla Foundation
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
import { Dict, Name, Ref, RefSetCache } from "../../src/core/primitives.js";
import { incrementalUpdate, writeDict } from "../../src/core/writer.js";
import { bytesToString } from "../../src/shared/util.js";
import { StringStream } from "../../src/core/stream.js";
describe("Writer", function () {
beforeAll(function () {
jasmine.clock().install();
jasmine.clock().mockDate(new Date(0));
});
afterAll(function () {
jasmine.clock().uninstall();
});
describe("Incremental update", function () {
it("should update a file with new objects", async function () {
const originalData = new Uint8Array();
const changes = new RefSetCache();
changes.put(Ref.get(123, 0x2d), { data: "abc\n" });
changes.put(Ref.get(456, 0x4e), { data: "defg\n" });
const xrefInfo = {
newRef: Ref.get(789, 0),
startXRef: 314,
fileIds: ["id", ""],
rootRef: null,
infoRef: null,
encryptRef: null,
filename: "foo.pdf",
infoMap: new Map(),
};
let data = await incrementalUpdate({
originalData,
xrefInfo,
changes,
xref: {},
useXrefStream: true,
});
data = bytesToString(data);
let expected =
"\nabc\n" +
"defg\n" +
"789 0 obj\n" +
"<< /Prev 314 /Size 790 /Type /XRef /Index [123 1 456 1 789 1] " +
"/W [1 1 1] /ID [(id) (\xeb\x4b\x2a\xe7\x31\x36\xf0\xcd\x83\x35\x94\x2a\x36\xcf\xaa\xb0)] " +
"/Length 9>> stream\n" +
"\x01\x01\x2d" +
"\x01\x05\x4e" +
"\x01\x0a\x00\n" +
"endstream\n" +
"endobj\n" +
"startxref\n" +
"10\n" +
"%%EOF\n";
expect(data).toEqual(expected);
data = await incrementalUpdate({
originalData,
xrefInfo,
changes,
xref: {},
useXrefStream: false,
});
data = bytesToString(data);
expected =
"\nabc\n" +
"defg\n" +
"xref\n" +
"123 1\n" +
"0000000001 00045 n\r\n" +
"456 1\n" +
"0000000005 00078 n\r\n" +
"789 1\n" +
"0000000010 00000 n\r\n" +
"trailer\n" +
"<< /Prev 314 /Size 789 " +
"/ID [(id) (\xeb\x4b\x2a\xe7\x31\x36\xf0\xcd\x83\x35\x94\x2a\x36\xcf\xaa\xb0)]>>\n" +
"startxref\n" +
"10\n" +
"%%EOF\n";
expect(data).toEqual(expected);
});
it("should update a file, missing the /ID-entry, with new objects", async function () {
const originalData = new Uint8Array();
const changes = new RefSetCache();
changes.put(Ref.get(123, 0x2d), { data: "abc\n" });
const xrefInfo = {
newRef: Ref.get(789, 0),
startXRef: 314,
fileIds: null,
rootRef: null,
infoRef: null,
encryptRef: null,
filename: "foo.pdf",
infoMap: new Map(),
};
let data = await incrementalUpdate({
originalData,
xrefInfo,
changes,
xref: {},
useXrefStream: true,
});
data = bytesToString(data);
const expected =
"\nabc\n" +
"789 0 obj\n" +
"<< /Prev 314 /Size 790 /Type /XRef /Index [123 1 789 1] " +
"/W [1 1 1] /Length 6>> stream\n" +
"\x01\x01\x2d" +
"\x01\x05\x00\n" +
"endstream\n" +
"endobj\n" +
"startxref\n" +
"5\n" +
"%%EOF\n";
expect(data).toEqual(expected);
});
});
describe("writeDict", function () {
it("should write a Dict", async function () {
const dict = new Dict(null);
dict.set("A", Name.get("B"));
dict.set("B", Ref.get(123, 456));
dict.set("C", 789);
dict.set("D", "hello world");
dict.set("E", "(hello\\world)");
dict.set("F", [1.23001, 4.50001, 6]);
const gdict = new Dict(null);
gdict.set("H", 123.00001);
const string = "a stream";
const stream = new StringStream(string);
stream.dict = new Dict(null);
stream.dict.set("Length", string.length);
gdict.set("I", stream);
dict.set("G", gdict);
dict.set("J", true);
dict.set("K", false);
dict.set("NullArr", [null, 10]);
dict.set("NullVal", null);
const buffer = [];
await writeDict(dict, buffer, null);
const expected =
"<< /A /B /B 123 456 R /C 789 /D (hello world) " +
"/E (\\(hello\\\\world\\)) /F [1.23001 4.50001 6] " +
"/G << /H 123.00001 /I << /Length 8>> stream\n" +
"a stream\n" +
"endstream>> /J true /K false " +
"/NullArr [null 10] /NullVal null>>";
expect(buffer.join("")).toEqual(expected);
});
it("should write a Dict in escaping PDF names", async function () {
const dict = new Dict(null);
dict.set("\xfeA#", Name.get("hello"));
dict.set("B", Name.get("#hello"));
dict.set("C", Name.get("he\xfello\xff"));
const buffer = [];
await writeDict(dict, buffer, null);
const expected = "<< /#feA#23 /hello /B /#23hello /C /he#fello#ff>>";
expect(buffer.join("")).toEqual(expected);
});
});
describe("XFA", function () {
it("should update AcroForm when no datasets in XFA array", async function () {
const originalData = new Uint8Array();
const changes = new RefSetCache();
const acroForm = new Dict(null);
acroForm.set("XFA", [
"preamble",
Ref.get(123, 0),
"postamble",
Ref.get(456, 0),
]);
const acroFormRef = Ref.get(789, 0);
const xfaDatasetsRef = Ref.get(101112, 0);
const xfaData = "<hello>world</hello>";
const xrefInfo = {
newRef: Ref.get(131415, 0),
startXRef: 314,
fileIds: null,
rootRef: null,
infoRef: null,
encryptRef: null,
filename: "foo.pdf",
infoMap: new Map(),
};
let data = await incrementalUpdate({
originalData,
xrefInfo,
changes,
hasXfa: true,
xfaDatasetsRef,
hasXfaDatasetsEntry: false,
acroFormRef,
acroForm,
xfaData,
xref: {},
useXrefStream: true,
});
data = bytesToString(data);
const expected =
"\n" +
"789 0 obj\n" +
"<< /XFA [(preamble) 123 0 R (datasets) 101112 0 R (postamble) 456 0 R]>>\n" +
"endobj\n" +
"101112 0 obj\n" +
"<< /Type /EmbeddedFile /Length 20>> stream\n" +
"<hello>world</hello>\n" +
"endstream\n" +
"endobj\n" +
"131415 0 obj\n" +
"<< /Prev 314 /Size 131416 /Type /XRef /Index [789 1 101112 1 131415 1] /W [1 1 0] /Length 6>> stream\n" +
"\x01\x01\x01[\x01¹\n" +
"endstream\n" +
"endobj\n" +
"startxref\n" +
"185\n" +
"%%EOF\n";
expect(data).toEqual(expected);
});
});
it("should update a file with a deleted object", async function () {
const originalData = new Uint8Array();
const changes = new RefSetCache();
changes.put(Ref.get(123, 0x2d), { data: null });
changes.put(Ref.get(456, 0x4e), { data: "abc\n" });
const xrefInfo = {
newRef: Ref.get(789, 0),
startXRef: 314,
fileIds: ["id", ""],
rootRef: null,
infoRef: null,
encryptRef: null,
filename: "foo.pdf",
infoMap: new Map(),
};
let data = await incrementalUpdate({
originalData,
xrefInfo,
changes,
xref: {},
useXrefStream: true,
});
data = bytesToString(data);
let expected =
"\nabc\n" +
"789 0 obj\n" +
"<< /Prev 314 /Size 790 /Type /XRef /Index [123 1 456 1 789 1] " +
"/W [1 1 1] /ID [(id) (\x5f\xd1\x43\x8e\xf8\x62\x79\x80\xbb\xd6\xf7\xb6\xd2\xb5\x6f\xd8)] " +
"/Length 9>> stream\n" +
"\x00\x00\x2e" +
"\x01\x01\x4e" +
"\x01\x05\x00\n" +
"endstream\n" +
"endobj\n" +
"startxref\n" +
"5\n" +
"%%EOF\n";
expect(data).toEqual(expected);
data = await incrementalUpdate({
originalData,
xrefInfo,
changes,
xref: {},
useXrefStream: false,
});
data = bytesToString(data);
expected =
"\nabc\n" +
"xref\n" +
"123 1\n" +
"0000000000 00046 f\r\n" +
"456 1\n" +
"0000000001 00078 n\r\n" +
"789 1\n" +
"0000000005 00000 n\r\n" +
"trailer\n" +
"<< /Prev 314 /Size 789 " +
"/ID [(id) (\x5f\xd1\x43\x8e\xf8\x62\x79\x80\xbb\xd6\xf7\xb6\xd2\xb5\x6f\xd8)]>>\n" +
"startxref\n" +
"5\n" +
"%%EOF\n";
expect(data).toEqual(expected);
});
});