Files
BadNote/lib/services/pptx_service.dart
Akiba So d346cc2670
All checks were successful
CI / Windows build (push) Successful in 14m22s
feat: unified shell, diagnostics pack, native Office, sticky board
Make Surface remote debugging and classroom workflows viable: always-on
structured logs with one-click zip export, a single AppShell chrome,
OOXML PPTX/DOCX annotation without LibreOffice, and a first-class sticky
board. Also drop spike/legacy ink widgets and tighten pen feel
(predictor, PenInfoHistory, page-tile layer).

Co-authored-by: Cursor <cursoragent@cursor.com>
2026-08-05 17:55:27 +08:00

365 lines
11 KiB
Dart

import 'dart:io';
import 'package:file_picker/file_picker.dart';
import 'package:path/path.dart' as p;
import 'package:path_provider/path_provider.dart';
import 'package:uuid/uuid.dart';
import 'office/office_document.dart';
import 'office/pptx_parser.dart';
/// Service for processing PPTX files: text extraction, structured slide parse,
/// optional LibreOffice image conversion, and file picking.
///
/// **Native OOXML parsing is primary** ([PptxParser] via `package:archive` +
/// `package:xml`). LibreOffice (`soffice`) is an optional fallback ONLY when
/// the native path fails AND the binary is present on the machine.
class PptxService {
static const _uuid = Uuid();
final PptxParser _parser;
PptxService({PptxParser? parser}) : _parser = parser ?? PptxParser();
/// Extract all text content from a PPTX file.
///
/// Prefers the native [PptxParser]. Falls back to a legacy unzip+regex path
/// only if native parsing throws.
Future<String> extractText(String pptxPath) async {
try {
final parsed = await _parser.parse(pptxPath);
return parsed.plainText;
} catch (_) {
return _extractTextLegacy(pptxPath);
}
}
/// Parse PPTX into structured slides (text runs with approximate positions,
/// embedded images extracted to a cache dir). Native-only — no LibreOffice.
Future<ParsedPptx> parseSlides(String pptxPath, {String? cacheDir}) {
return _parser.parse(pptxPath, cacheDirPath: cacheDir);
}
/// Convert PPTX slides to a list of image file paths (legacy PenSlideScreen).
///
/// Prefer [parseSlides] for native text+image rendering. This method only
/// invokes LibreOffice when native parse fails AND `soffice` exists;
/// otherwise it emits placeholder PNGs from the native slide count.
Future<List<String>> convertToImages(String pptxPath) async {
try {
final parsed = await _parser.parse(pptxPath);
// Native succeeded — do NOT call LibreOffice; placeholders for callers
// that still expect image paths. OfficeDocumentScreen uses [parseSlides].
if (parsed.slides.isEmpty) return [];
return _generatePlaceholderImages(pptxPath);
} catch (_) {
// Native failed — LibreOffice fallback ONLY if soffice exists.
final soffice = await resolveSoffice();
if (soffice != null) {
final loImages = await _convertViaLibreOffice(pptxPath);
if (loImages.isNotEmpty) return loImages;
}
return _generatePlaceholderImages(pptxPath);
}
}
/// Open a file picker dialog and return the selected PPTX path, or null.
Future<String?> openPptxFile() async {
final result = await FilePicker.platform.pickFiles(
type: FileType.custom,
allowedExtensions: ['pptx', 'ppt'],
);
final files = result?.files;
if (files == null || files.isEmpty) return null;
return files.first.path;
}
// ---------------------------------------------------------------------------
// Legacy / LibreOffice helpers
// ---------------------------------------------------------------------------
Future<String> _extractTextLegacy(String pptxPath) async {
final tmpDir = await _makeTmpDir('pptx_text');
try {
final unzipResult = await Process.run('unzip', [
'-o',
'-q',
pptxPath,
'-d',
tmpDir.path,
]);
if (unzipResult.exitCode != 0) {
return '';
}
final slidesDir = Directory(p.join(tmpDir.path, 'ppt', 'slides'));
if (!await slidesDir.exists()) return '';
final slideFiles = await slidesDir
.list()
.where((f) => f.path.contains(RegExp(r'slide\d+\.xml$')))
.toList();
slideFiles.sort((a, b) {
final aNum = _extractSlideNumber(a.path);
final bNum = _extractSlideNumber(b.path);
return aNum.compareTo(bNum);
});
final buffer = StringBuffer();
for (final slideFile in slideFiles) {
final xml = await File(slideFile.path).readAsString();
final slideText = _extractTextFromXml(xml);
if (slideText.isNotEmpty) {
final num = _extractSlideNumber(slideFile.path);
buffer.writeln('--- Slide $num ---');
buffer.writeln(slideText);
buffer.writeln();
}
}
return buffer.toString().trim();
} catch (_) {
return '';
} finally {
try {
await tmpDir.delete(recursive: true);
} catch (_) {}
}
}
String _extractTextFromXml(String xml) {
final lines = <String>[];
final regex = RegExp(r'<a:t[^>]*>(.*?)</a:t>', dotAll: true);
for (final match in regex.allMatches(xml)) {
final text = match.group(1) ?? '';
if (text.trim().isNotEmpty) {
lines.add(text.trim());
}
}
return lines.join('\n');
}
int _extractSlideNumber(String path) {
final match = RegExp(r'slide(\d+)\.xml$').firstMatch(path);
if (match != null) return int.parse(match.group(1)!);
return 0;
}
/// LibreOffice fallback — only when native fails or caller wants PNGs and
/// soffice is installed.
Future<List<String>> _convertViaLibreOffice(String pptxPath) async {
try {
final soffice = await resolveSoffice();
if (soffice == null) return [];
final outDir = await _makeTmpDir('pptx_images');
final result = await Process.run(soffice, [
'--headless',
'--convert-to',
'png',
'--outdir',
outDir.path,
pptxPath,
]);
if (result.exitCode != 0) return [];
final pngs = await outDir
.list()
.where((f) => f.path.endsWith('.png'))
.map((f) => f.path)
.toList();
pngs.sort();
final persistDir = await _makeTmpDir('pptx_slides');
final persistentPaths = <String>[];
for (var i = 0; i < pngs.length; i++) {
final src = File(pngs[i]);
final dst = p.join(persistDir.path, 'slide_${i + 1}.png');
await src.copy(dst);
persistentPaths.add(dst);
}
try {
await outDir.delete(recursive: true);
} catch (_) {}
return persistentPaths;
} catch (_) {
return [];
}
}
/// Resolve the LibreOffice CLI binary, or null when unavailable.
static Future<String?> resolveSoffice() async {
if (Platform.isWindows) {
const candidates = [
r'C:\Program Files\LibreOffice\program\soffice.exe',
r'C:\Program Files (x86)\LibreOffice\program\soffice.exe',
];
for (final c in candidates) {
if (await File(c).exists()) return c;
}
if (await _whichOk('soffice')) return 'soffice';
return null;
}
if (await _whichOk('libreoffice')) return 'libreoffice';
if (await _whichOk('soffice')) return 'soffice';
return null;
}
static Future<bool> _whichOk(String cmd) async {
try {
final r = await Process.run('which', [cmd]);
return r.exitCode == 0;
} catch (_) {
return false;
}
}
/// Convert an arbitrary office document (e.g. DOCX) to PDF via LibreOffice.
/// Optional — native [DocxParser] is preferred for opening in BadNote.
Future<String?> convertToPdf(String sourcePath) async {
final soffice = await resolveSoffice();
if (soffice == null) return null;
final outDir = p.dirname(sourcePath);
try {
final result = await Process.run(soffice, [
'--headless',
'--convert-to',
'pdf',
'--outdir',
outDir,
sourcePath,
]);
if (result.exitCode != 0) return null;
final pdfPath = p.join(
outDir,
'${p.basenameWithoutExtension(sourcePath)}.pdf',
);
if (await File(pdfPath).exists()) return pdfPath;
return null;
} catch (_) {
return null;
}
}
Future<List<String>> _generatePlaceholderImages(String pptxPath) async {
final slideCount = await _countSlidesNative(pptxPath);
if (slideCount == 0) return [];
final outDir = await _makeTmpDir('pptx_placeholders');
final paths = <String>[];
final hasConvert = await _hasCommand('convert');
for (var i = 1; i <= slideCount; i++) {
final path = p.join(outDir.path, 'slide_$i.png');
if (hasConvert) {
await _generateWithImageMagick(path, i, slideCount);
} else {
await _writeMinimalPng(path);
}
paths.add(path);
}
return paths;
}
Future<int> _countSlidesNative(String pptxPath) async {
try {
final parsed = await _parser.parse(pptxPath);
return parsed.slides.length;
} catch (_) {
return _countSlidesUnzip(pptxPath);
}
}
Future<int> _countSlidesUnzip(String pptxPath) async {
final tmpDir = await _makeTmpDir('pptx_count');
try {
await Process.run('unzip', ['-o', '-q', pptxPath, '-d', tmpDir.path]);
final slidesDir = Directory(p.join(tmpDir.path, 'ppt', 'slides'));
if (!await slidesDir.exists()) return 0;
final count = await slidesDir
.list()
.where((f) => f.path.contains(RegExp(r'slide\d+\.xml$')))
.length;
return count;
} catch (_) {
return 0;
} finally {
try {
await tmpDir.delete(recursive: true);
} catch (_) {}
}
}
Future<bool> _hasCommand(String cmd) async {
try {
final result = await Process.run('which', [cmd]);
return result.exitCode == 0;
} catch (_) {
return false;
}
}
Future<void> _generateWithImageMagick(
String outPath,
int slideNum,
int total,
) async {
final hue = ((slideNum - 1) * 137) % 360;
await Process.run('convert', [
'-size',
'1920x1080',
'xc:hsl($hue, 60%, 92%)',
'-gravity',
'center',
'-pointsize',
'120',
'-fill',
'hsl($hue, 30%, 40%)',
'-annotate',
'+0+0',
'Slide $slideNum / $total',
outPath,
]);
}
Future<void> _writeMinimalPng(String path) async {
const pngBytes = <int>[
0x89, 0x50, 0x4E, 0x47, 0x0D, 0x0A, 0x1A, 0x0A,
0x00, 0x00, 0x00, 0x0D,
0x49, 0x48, 0x44, 0x52,
0x00, 0x00, 0x00, 0x01,
0x00, 0x00, 0x00, 0x01,
0x08, 0x02,
0x00, 0x00, 0x00,
0x90, 0x77, 0x53, 0xDE,
0x00, 0x00, 0x00, 0x0C,
0x49, 0x44, 0x41, 0x54,
0x08, 0xD7, 0x63, 0xF8, 0xCF, 0xC0, 0x00, 0x00,
0x01, 0x01, 0x01, 0x00,
0x18, 0xDD, 0x8D, 0xB4,
0x00, 0x00, 0x00, 0x00,
0x49, 0x45, 0x4E, 0x44,
0xAE, 0x42, 0x60, 0x82,
];
await File(path).writeAsBytes(pngBytes);
}
Future<Directory> _makeTmpDir(String prefix) async {
final base = await getTemporaryDirectory();
final dir = Directory(p.join(base.path, '${prefix}_${_uuid.v4()}'));
await dir.create(recursive: true);
return dir;
}
}