clank/clank-together
1
0
Fork 0
Code Issues Wiki Activity

Validate original literals using the configured execution charset

Browse source
This commit is contained in:
ragnarok-clanker 2026-10-27 10:45:49 +01:00
commit b20e903244
10 changed files with 150 additions and 33 deletions
Download patch file Download diff file Expand all files Collapse all files

25
test/freetext.test.js
View file

@ -1,6 +1,6 @@
import { test } from 'node:test';
import assert from 'node:assert/strict';
import { scanText, freeTextErrors, freeTextErrorsAcross, allowedComment } from '../src/shared/freetext.js';
import { scanText, freeTextErrors, freeTextErrorsAcross, allowedComment, executionCharset } from '../src/shared/freetext.js';
// Stands in for the original executable's bytes.
const original = Buffer.concat([Buffer.from('Interface\\GlueXML\0', 'latin1'), Buffer.from('hello world\0%d\0', 'latin1'), Buffer.from('Wide\0', 'utf16le')]);
@ -63,3 +63,26 @@ test('pieces of a concatenated literal may each come from the original', () => {
assert.deepEqual(freeTextErrors('const char* s = "%d" "hello world";\n', null, inOriginal), []);
assert.equal(freeTextErrors('const char* s = "%d" "made up";\n', null, inOriginal).length, 1);
});
test('execution charset matches Hangul and universal escapes while preserving numeric byte escapes', () => {
const original = Buffer.from('c0afc0fa005769646500', 'hex');
const found = (bytes) => original.includes(bytes);
assert.equal(executionCharset(['/source-charset:utf-8', '/execution-charset:.949']), 'cp949');
assert.equal(executionCharset(['/execution-charset:.949', '/utf-8']), 'utf8');
assert.deepEqual(freeTextErrors('const char* s = "유저";', null, found, 'cp949'), []);
assert.deepEqual(freeTextErrors(String.raw`const char* s = "\uC720\uC800";`, null, found, 'cp949'), []);
assert.deepEqual(freeTextErrors(String.raw`const char* s = "\xC0\xAF\xC0\xFA";`, null, found, 'cp949'), []);
assert.deepEqual(freeTextErrors(String.raw`const char* s = "유\xC0\xFA";`, null, found, 'cp949'), []);
assert.equal(freeTextErrors('const char* s = "유저";', null, found, 'utf8').length, 1);
assert.equal(freeTextErrors('const char* s = "새 문자열";', null, found, 'cp949').length, 1);
assert.equal(freeTextErrors('const char* s = "😀";', null, (b) => b.equals(Buffer.from('?')), 'cp949').length, 1, 'unrepresentable Unicode cannot pass through replacement bytes');
assert.deepEqual(scanText('const wchar_t* s = L"유저";', 'cp949').strings[0].pieces, [Buffer.from('유저', 'utf16le')]);
});
test('mixed execution charsets preserve pooled original byte identity across moves', () => {
const baseFiles = { 'a.cpp': String.raw`const char* s = "\xC0\xAF\xC0\xFA";`, 'b.cpp': null };
const options = { baseFiles, charsetForFile: (file) => file === 'a.cpp' ? 'utf8' : 'cp949' };
assert.deepEqual(freeTextErrorsAcross({ 'a.cpp': '', 'b.cpp': 'const char* s = "유저";' }, '', () => false, options), []);
const twice = freeTextErrorsAcross({ 'a.cpp': baseFiles['a.cpp'], 'b.cpp': 'const char* s = "유저";' }, '', () => false, options);
assert.equal(twice.length, 1, 'one original occurrence does not exempt an extra copy');
});