LWLT-AIBOT/control-plane/test/document-converter.test.ts

import assert from 'node:assert/strict';
import { chmod, mkdtemp, rm, writeFile } from 'node:fs/promises';
import { join } from 'node:path';
import { tmpdir } from 'node:os';
import test from 'node:test';
import { loadConfig } from '../src/config.js';
import { convertDocumentToPdf, convertDocumentToXlsx } from '../src/document-converter.js';

function testConfig(converterPath: string, fontConfigFile?: string) {
  return loadConfig({
    NODE_ENV: 'test',
    FIELD_ENCRYPTION_KEY: Buffer.alloc(32, 21).toString('base64'),
    DOCUMENT_CONVERTER_PATH: converterPath,
    ...(fontConfigFile ? { DOCUMENT_FONTCONFIG_FILE: fontConfigFile } : {})
  });
}

test('PDF sources pass through with PDF metadata', async () => {
  const content = Buffer.from('%PDF-1.7\nsource');
  const result = await convertDocumentToPdf(testConfig('/path/that/is/not/used'), {
    fileName: 'team.doc',
    contentType: 'application/msword',
    content,
    maxBytes: 10_000
  });
  assert.equal(result.fileName, 'team.pdf');
  assert.equal(result.contentType, 'application/pdf');
  assert.equal(result.converted, true);
  assert.equal(result.conversion_status, 'already_pdf');
  assert.deepEqual(result.content, content);
});

test('conversion failure falls back to the source file', async () => {
  const content = Buffer.from('not a valid office document');
  const result = await convertDocumentToPdf(testConfig('/path/that/does/not/exist'), {
    fileName: 'team.doc',
    contentType: 'application/msword',
    content,
    maxBytes: 10_000
  });
  assert.equal(result.fileName, 'team.doc');
  assert.equal(result.contentType, 'application/msword');
  assert.equal(result.converted, false);
  assert.equal(result.conversion_status, 'source_fallback');
  assert.equal(result.conversion_error_code, 'converter_unavailable');
  assert.deepEqual(result.content, content);
});

test('a valid converter output replaces the source with a PDF', {
  skip: process.platform === 'win32' ? 'Unix fake converter fixture is not executable through Windows execFile.' : false
}, async () => {
  const temporaryDirectory = await mkdtemp(join(tmpdir(), 'ltjt-converter-test-'));
  const converterPath = join(temporaryDirectory, 'fake-soffice.sh');
  const fontConfigFile = join(temporaryDirectory, 'fonts.conf');
  await writeFile(fontConfigFile, '<fontconfig/>');
  await writeFile(converterPath, [
    '#!/bin/sh',
    'outdir=""',
    'convert_to=""',
    'next_is_outdir=0',
    'next_is_convert_to=0',
    'for arg in "$@"; do',
    '  if [ "$next_is_convert_to" = "1" ]; then convert_to="$arg"; next_is_convert_to=0; continue; fi',
    '  if [ "$next_is_outdir" = "1" ]; then outdir="$arg"; next_is_outdir=0; continue; fi',
    '  if [ "$arg" = "--convert-to" ]; then next_is_convert_to=1; fi',
    '  if [ "$arg" = "--outdir" ]; then next_is_outdir=1; fi',
    'done',
    '[ "$convert_to" = "pdf:writer_pdf_Export" ] || exit 41',
    '[ "$FONTCONFIG_FILE" = "' + fontConfigFile + '" ] || exit 42',
    'printf "%s" "%PDF-1.7\\nconverted" > "$outdir/team.pdf"'
  ].join('\n'), { mode: 0o700 });
  await chmod(converterPath, 0o700);
  try {
    const result = await convertDocumentToPdf(testConfig(converterPath, fontConfigFile), {
      fileName: 'team.docx',
      contentType: 'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
      content: Buffer.from('source'),
      maxBytes: 10_000
    });
    assert.equal(result.fileName, 'team.pdf');
    assert.equal(result.contentType, 'application/pdf');
    assert.equal(result.converted, true);
    assert.equal(result.conversion_status, 'converted');
    assert.match(result.content.toString(), /^%PDF-1\.7/);
  } finally {
    await rm(temporaryDirectory, { recursive: true, force: true });
  }
});

test('HTML-in-XLS visitor sources become real XLSX workbooks', {
  skip: process.platform === 'win32' ? 'Unix fake converter fixture is not executable through Windows execFile.' : false
}, async () => {
  const temporaryDirectory = await mkdtemp(join(tmpdir(), 'ltjt-converter-xlsx-test-'));
  const converterPath = join(temporaryDirectory, 'fake-soffice.sh');
  await writeFile(converterPath, [
    '#!/bin/sh',
    'outdir=""',
    'convert_to=""',
    'input_path=""',
    'next_is_outdir=0',
    'next_is_convert_to=0',
    'for arg in "$@"; do',
    '  if [ "$next_is_convert_to" = "1" ]; then convert_to="$arg"; next_is_convert_to=0; continue; fi',
    '  if [ "$next_is_outdir" = "1" ]; then outdir="$arg"; next_is_outdir=0; continue; fi',
    '  if [ "$arg" = "--convert-to" ]; then next_is_convert_to=1; continue; fi',
    '  if [ "$arg" = "--outdir" ]; then next_is_outdir=1; continue; fi',
    '  case "$arg" in *.xls) input_path="$arg";; esac',
    'done',
    '[ "$convert_to" = "xlsx:Calc MS Excel 2007 XML" ] || exit 41',
    'case "$input_path" in *.xls) ;; *) exit 42;; esac',
    'printf "PK\\003\\004fake-xlsx" > "$outdir/Visitor.xlsx"'
  ].join('\n'), { mode: 0o700 });
  await chmod(converterPath, 0o700);
  try {
    const content = Buffer.from('<!doctype html><html><body><table><tr><td>游客</td></tr></table></body></html>', 'utf8');
    const result = await convertDocumentToXlsx(testConfig(converterPath), {
      fileName: 'Visitor.xls',
      contentType: 'application/vnd.ms-excel; Charset=UTF-8',
      content,
      maxBytes: 10_000
    });
    assert.equal(result.fileName, 'Visitor.xlsx');
    assert.equal(result.contentType, 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet');
    assert.equal(result.converted, true);
    assert.equal(result.conversion_status, 'converted_xlsx');
    assert.deepEqual(result.content.subarray(0, 4), Buffer.from([0x50, 0x4b, 0x03, 0x04]));
  } finally {
    await rm(temporaryDirectory, { recursive: true, force: true });
  }
});

test('visitor XLSX conversion failure remains a source fallback', async () => {
  const content = Buffer.from('<html><body>游客</body></html>', 'utf8');
  const result = await convertDocumentToXlsx(testConfig('/path/that/does/not/exist'), {
    fileName: 'Visitor.xls',
    contentType: 'application/vnd.ms-excel',
    content,
    maxBytes: 10_000
  });
  assert.equal(result.fileName, 'Visitor.xls');
  assert.equal(result.converted, false);
  assert.equal(result.conversion_status, 'source_fallback');
  assert.equal(result.conversion_error_code, 'converter_unavailable');
  assert.deepEqual(result.content, content);
});

test('wide Word HTML tables are scaled to the printable width before conversion', {
  skip: process.platform === 'win32' ? 'Unix fake converter fixture is not executable through Windows execFile.' : false
}, async () => {
  const temporaryDirectory = await mkdtemp(join(tmpdir(), 'ltjt-converter-html-test-'));
  const converterPath = join(temporaryDirectory, 'fake-soffice.sh');
  await writeFile(converterPath, [
    '#!/bin/sh',
    'outdir=""',
    'input_path=""',
    'next_is_outdir=0',
    'for arg in "$@"; do',
    '  if [ "$next_is_outdir" = "1" ]; then outdir="$arg"; next_is_outdir=0; continue; fi',
    '  if [ "$arg" = "--outdir" ]; then next_is_outdir=1; continue; fi',
    '  case "$arg" in *.doc|*.docx|*.html) input_path="$arg";; esac',
    'done',
    'grep -q "width=733" "$input_path" || exit 43',
    'grep -q "width=\\\"154\\\"" "$input_path" || exit 44',
    'grep -q "width: 550pt" "$input_path" || exit 45',
    'printf "%s" "%PDF-1.7\\nconverted" > "$outdir/team.pdf"'
  ].join('\n'), { mode: 0o700 });
  await chmod(converterPath, 0o700);
  try {
    const result = await convertDocumentToPdf(testConfig(converterPath), {
      fileName: 'team.doc',
      contentType: 'application/msword',
      content: Buffer.from(
        '<html><head><style>table { width: 760px; } .inner { width: 570pt; }</style></head>'
          + '<body><table width=760><tr><td width="160" class="inner">内容</td></tr></table></body></html>',
        'utf8'
      ),
      maxBytes: 10_000
    });
    assert.equal(result.converted, true);
    assert.equal(result.conversion_status, 'converted');
  } finally {
    await rm(temporaryDirectory, { recursive: true, force: true });
  }
});