ktg-plugin-marketplace/plugins/llm-security/tests/scanners/zip-extract.test.mjs

// zip-extract.test.mjs — Unit tests for the zero-dep ZIP extractor.

import { describe, it } from 'node:test';
import assert from 'node:assert/strict';
import { mkdtemp, rm, readFile, readdir } from 'node:fs/promises';
import { tmpdir } from 'node:os';
import { join } from 'node:path';
import { deflateRawSync } from 'node:zlib';
import { extractToDir, listEntries, ZipError, __testing } from '../../scanners/lib/zip-extract.mjs';
import { buildZip, unixModeAttr, MODE_SYMLINK } from '../lib/build-zip.mjs';

const { validateEntryName, isSymlink, DEFAULT_CAPS } = __testing;

async function withTempDir(fn) {
  const dir = await mkdtemp(join(tmpdir(), 'zip-test-'));
  try { return await fn(dir); }
  finally { await rm(dir, { recursive: true, force: true }); }
}

describe('validateEntryName', () => {
  it('accepts a normal nested path', () => {
    const out = validateEntryName('extension/package.json', DEFAULT_CAPS);
    assert.ok(out && out.includes('package.json'));
  });
  it('returns null for directory entries', () => {
    assert.equal(validateEntryName('extension/', DEFAULT_CAPS), null);
  });
  it('rejects parent traversal', () => {
    assert.throws(() => validateEntryName('../etc/passwd', DEFAULT_CAPS), /traversal/);
  });
  it('rejects deep parent traversal', () => {
    assert.throws(() => validateEntryName('extension/../../escape', DEFAULT_CAPS), /traversal/);
  });
  it('rejects POSIX absolute paths', () => {
    assert.throws(() => validateEntryName('/etc/passwd', DEFAULT_CAPS), /absolute/);
  });
  it('rejects Windows drive letters', () => {
    assert.throws(() => validateEntryName('C:\\Windows\\sys', DEFAULT_CAPS), /drive-letter|absolute/);
  });
  it('rejects backslash absolute paths', () => {
    assert.throws(() => validateEntryName('\\foo', DEFAULT_CAPS), /absolute/);
  });
  it('rejects NUL bytes', () => {
    assert.throws(() => validateEntryName('foo\u0000bar', DEFAULT_CAPS), /NUL/);
  });
  it('rejects empty entry names', () => {
    assert.throws(() => validateEntryName('', DEFAULT_CAPS), /empty/);
  });
  it('rejects very deep paths beyond depth cap', () => {
    const deep = Array.from({ length: 25 }, () => 'a').join('/');
    assert.throws(() => validateEntryName(deep, { ...DEFAULT_CAPS, maxDepth: 20 }), /depth/);
  });
  it('normalizes backslashes in path', () => {
    const out = validateEntryName('extension\\sub\\file.txt', DEFAULT_CAPS);
    assert.ok(out && (out.includes('sub') || out.includes('file.txt')));
  });
});

describe('isSymlink', () => {
  it('detects unix-made symlink mode bits', () => {
    const entry = { versionMadeBy: (3 << 8) | 20, externalAttr: unixModeAttr(MODE_SYMLINK) };
    assert.equal(isSymlink(entry), true);
  });
  it('ignores mode bits when versionMadeBy os != Unix', () => {
    const entry = { versionMadeBy: (0 << 8) | 20, externalAttr: unixModeAttr(MODE_SYMLINK) };
    assert.equal(isSymlink(entry), false);
  });
  it('returns false for regular file', () => {
    const entry = { versionMadeBy: (3 << 8) | 20, externalAttr: unixModeAttr(0x81A4) };
    assert.equal(isSymlink(entry), false);
  });
});

describe('extractToDir — happy path', () => {
  it('extracts a small ZIP with a nested file', async () => {
    const buf = buildZip([
      { name: 'extension/package.json', data: '{"hello":"world"}' },
      { name: 'extension/extension.js', data: 'console.log(1)' },
    ]);
    await withTempDir(async (dir) => {
      const r = await extractToDir(buf, dir);
      assert.equal(r.entries, 2);
      const pkg = await readFile(join(dir, 'extension/package.json'), 'utf8');
      assert.match(pkg, /hello/);
    });
  });

  it('extracts deflate-compressed entries', async () => {
    // Pseudo-random bytes so compression ratio stays well under the cap.
    const original = Buffer.alloc(2000);
    for (let i = 0; i < original.length; i++) original[i] = (i * 73 + 11) & 0xFF;
    const compressed = deflateRawSync(original);
    // Manually construct a buildZip-style entry but with method=8 + compSize set.
    // buildZip only supports STORE; we need a small bespoke builder for this test.
    // Use raw buildZip + override method by patching after.
    // Simpler: assert listEntries handles a deflate one we craft.
    // Construct manually:
    const nameBuf = Buffer.from('extension/big.txt', 'utf8');
    const lfh = Buffer.alloc(30);
    lfh.writeUInt32LE(0x04034b50, 0);
    lfh.writeUInt16LE(20, 4);
    lfh.writeUInt16LE(0, 6);
    lfh.writeUInt16LE(8, 8); // DEFLATE
    lfh.writeUInt32LE(0, 14); // CRC unused (we don't validate)
    lfh.writeUInt32LE(compressed.length, 18);
    lfh.writeUInt32LE(original.length, 22);
    lfh.writeUInt16LE(nameBuf.length, 26);
    lfh.writeUInt16LE(0, 28);
    const cd = Buffer.alloc(46);
    cd.writeUInt32LE(0x02014b50, 0);
    cd.writeUInt16LE(20, 4);
    cd.writeUInt16LE(20, 6);
    cd.writeUInt16LE(0, 8);
    cd.writeUInt16LE(8, 10); // DEFLATE
    cd.writeUInt32LE(0, 16);
    cd.writeUInt32LE(compressed.length, 20);
    cd.writeUInt32LE(original.length, 24);
    cd.writeUInt16LE(nameBuf.length, 28);
    cd.writeUInt32LE(0, 38);
    cd.writeUInt32LE(0, 42); // LFH at offset 0
    const eocd = Buffer.alloc(22);
    eocd.writeUInt32LE(0x06054b50, 0);
    eocd.writeUInt16LE(1, 8);
    eocd.writeUInt16LE(1, 10);
    eocd.writeUInt32LE(46 + nameBuf.length, 12);
    eocd.writeUInt32LE(30 + nameBuf.length + compressed.length, 16);
    const buf = Buffer.concat([lfh, nameBuf, compressed, cd, nameBuf, eocd]);

    await withTempDir(async (dir) => {
      const r = await extractToDir(buf, dir);
      assert.equal(r.entries, 1);
      const out = await readFile(join(dir, 'extension/big.txt'));
      assert.equal(out.length, original.length);
      assert.equal(out.toString('utf8'), original.toString('utf8'));
    });
  });

  it('lists entries without extracting', () => {
    const buf = buildZip([{ name: 'a.txt', data: 'x' }, { name: 'b.txt', data: 'yy' }]);
    const out = listEntries(buf);
    assert.equal(out.length, 2);
    assert.equal(out[0].name, 'a.txt');
    assert.equal(out[1].uncompSize, 2);
  });
});

describe('extractToDir — adversarial', () => {
  it('rejects zip-slip via parent traversal', async () => {
    const buf = buildZip([{ name: '../escape.txt', data: 'pwned' }]);
    await withTempDir(async (dir) => {
      await assert.rejects(() => extractToDir(buf, dir), /traversal/);
      const items = await readdir(dir);
      assert.equal(items.length, 0, 'no files should have been written');
    });
  });

  it('rejects zip-slip via absolute POSIX path', async () => {
    const buf = buildZip([{ name: '/tmp/leak.txt', data: 'pwned' }]);
    await withTempDir(async (dir) => {
      await assert.rejects(() => extractToDir(buf, dir), /absolute|traversal/);
    });
  });

  it('rejects symlink entries', async () => {
    const buf = buildZip([{
      name: 'evil-link',
      data: '../../etc/passwd',
      versionMadeBy: (3 << 8) | 20,
      externalAttr: unixModeAttr(MODE_SYMLINK),
    }]);
    await withTempDir(async (dir) => {
      await assert.rejects(() => extractToDir(buf, dir), /symlink/);
    });
  });

  it('rejects entries beyond maxEntries cap', async () => {
    const entries = Array.from({ length: 5 }, (_, i) => ({ name: `f${i}.txt`, data: 'x' }));
    const buf = buildZip(entries);
    await withTempDir(async (dir) => {
      await assert.rejects(
        () => extractToDir(buf, dir, { caps: { ...DEFAULT_CAPS, maxEntries: 3 } }),
        /too many/,
      );
    });
  });

  it('rejects zip-bomb: STORED entry exceeding maxUncompressedBytes', async () => {
    const buf = buildZip([{
      name: 'bomb.txt',
      data: Buffer.alloc(2000),
      declaredUncompSize: 2000,
    }]);
    await withTempDir(async (dir) => {
      await assert.rejects(
        () => extractToDir(buf, dir, { caps: { ...DEFAULT_CAPS, maxUncompressedBytes: 1000 } }),
        /maxUncompressedBytes/,
      );
    });
  });

  it('rejects zip-bomb: deflate expansion ratio exceeds cap', async () => {
    // Build an entry with high uncompressed and tiny compressed claim.
    const original = Buffer.alloc(20_000); // 20KB of zeros — compresses tiny
    const compressed = deflateRawSync(original);
    const nameBuf = Buffer.from('bomb.bin', 'utf8');
    const lfh = Buffer.alloc(30);
    lfh.writeUInt32LE(0x04034b50, 0);
    lfh.writeUInt16LE(20, 4);
    lfh.writeUInt16LE(8, 8);
    lfh.writeUInt32LE(compressed.length, 18);
    lfh.writeUInt32LE(original.length, 22);
    lfh.writeUInt16LE(nameBuf.length, 26);
    const cd = Buffer.alloc(46);
    cd.writeUInt32LE(0x02014b50, 0);
    cd.writeUInt16LE(20, 4); cd.writeUInt16LE(20, 6); cd.writeUInt16LE(8, 10);
    cd.writeUInt32LE(compressed.length, 20);
    cd.writeUInt32LE(original.length, 24);
    cd.writeUInt16LE(nameBuf.length, 28);
    cd.writeUInt32LE(0, 42);
    const eocd = Buffer.alloc(22);
    eocd.writeUInt32LE(0x06054b50, 0);
    eocd.writeUInt16LE(1, 8); eocd.writeUInt16LE(1, 10);
    eocd.writeUInt32LE(46 + nameBuf.length, 12);
    eocd.writeUInt32LE(30 + nameBuf.length + compressed.length, 16);
    const buf = Buffer.concat([lfh, nameBuf, compressed, cd, nameBuf, eocd]);
    await withTempDir(async (dir) => {
      await assert.rejects(
        () => extractToDir(buf, dir, { caps: { ...DEFAULT_CAPS, maxExpansionRatio: 5 } }),
        /expansion ratio|exceeds/,
      );
    });
  });

  it('rejects unknown compression methods', async () => {
    // Manually craft an entry with method=6 (Implode, unsupported)
    const nameBuf = Buffer.from('weird.bin', 'utf8');
    const data = Buffer.from('x');
    const lfh = Buffer.alloc(30);
    lfh.writeUInt32LE(0x04034b50, 0);
    lfh.writeUInt16LE(6, 8); // method=Implode
    lfh.writeUInt32LE(data.length, 18);
    lfh.writeUInt32LE(data.length, 22);
    lfh.writeUInt16LE(nameBuf.length, 26);
    const cd = Buffer.alloc(46);
    cd.writeUInt32LE(0x02014b50, 0);
    cd.writeUInt16LE(6, 10);
    cd.writeUInt32LE(data.length, 20);
    cd.writeUInt32LE(data.length, 24);
    cd.writeUInt16LE(nameBuf.length, 28);
    const eocd = Buffer.alloc(22);
    eocd.writeUInt32LE(0x06054b50, 0);
    eocd.writeUInt16LE(1, 8); eocd.writeUInt16LE(1, 10);
    eocd.writeUInt32LE(46 + nameBuf.length, 12);
    eocd.writeUInt32LE(30 + nameBuf.length + data.length, 16);
    const buf = Buffer.concat([lfh, nameBuf, data, cd, nameBuf, eocd]);
    await withTempDir(async (dir) => {
      await assert.rejects(() => extractToDir(buf, dir), /unsupported compression/);
    });
  });

  it('throws ZipError when EOCD is missing', async () => {
    const garbage = Buffer.from('not a zip file at all');
    await withTempDir(async (dir) => {
      await assert.rejects(() => extractToDir(garbage, dir), /EOCD/);
    });
  });
});