Bug fixes (Flutter): - Wrap multi-statement DB writes (insert/update/delete note, deleteDocument, deletePageData, OCR FTS merge, migrations) in transactions to prevent data loss on interruption and a read-modify-write FTS race. - Fix PdfDocument leaks on exception (try/finally dispose) and preserve image aspect ratio when stamping images onto PDF pages. - Guard file-picker against empty selection (was .single -> crash). - Fix eraser ConcurrentModificationError and unmodifiable-list crash on PDF pages; capture page synchronously on save to stop wrong-page data loss. - Fix Riverpod DB-not-ready races, broken pull-to-refresh, settings load race, and search N+1; transform stored annotations on PDF page rotation. - Normalize pen pressure for devices without a pressure range. - PPT: single source of truth for slide strokes so ink displays and exports. UI/UX: - Material 3 typography, theme-aware colors (dark-mode fixes), hover cursors and right-click/visible actions on desktop, keyboard shortcuts (undo/redo/ save/find), toolbar overflow handling, friendlier empty states, semantic OCR status badges, relative timestamps, 1-based page indicators, large-deck PPT navigation, and a scratchpad-scope label in split view. Server (optional backend): - Persist JWT secret (was per-process random), block path traversal in storage, fix CORS '*'+credentials, add OCR job ownership checks, last-writer-wins sync guard, constant-time login, and split out heavy OCR deps so the API/tests run without them. CI: Gitea workflows for format+analyze+test (Linux, system sqlite) and a Windows release build; pristine `flutter analyze`, all Flutter and server tests green. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
294 lines
8.9 KiB
Dart
294 lines
8.9 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';
|
|
|
|
/// Service for processing PPTX files: text extraction, image conversion, file picking.
|
|
///
|
|
/// PPTX files are ZIP archives containing XML. We extract text from
|
|
/// `ppt/slides/slide*.xml` `<a:t>` elements and convert slides to images
|
|
/// using LibreOffice (headless) or generate placeholder images as fallback.
|
|
class PptxService {
|
|
static const _uuid = Uuid();
|
|
|
|
/// Extract all text content from a PPTX file.
|
|
///
|
|
/// PPTX is a ZIP archive. Slide text lives in `ppt/slides/slide*.xml`
|
|
/// inside `<a:t>` (ASCII text) elements within `<a:r>` (run) or
|
|
/// `<a:p>` (paragraph) nodes.
|
|
Future<String> extractText(String pptxPath) async {
|
|
final tmpDir = await _makeTmpDir('pptx_text');
|
|
|
|
try {
|
|
// Unzip the PPTX
|
|
final unzipResult = await Process.run('unzip', [
|
|
'-o',
|
|
'-q',
|
|
pptxPath,
|
|
'-d',
|
|
tmpDir.path,
|
|
]);
|
|
|
|
if (unzipResult.exitCode != 0) {
|
|
return '';
|
|
}
|
|
|
|
// Find all slide XML files
|
|
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();
|
|
|
|
// Sort by slide number
|
|
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 {
|
|
// Cleanup
|
|
try {
|
|
await tmpDir.delete(recursive: true);
|
|
} catch (_) {}
|
|
}
|
|
}
|
|
|
|
/// Convert PPTX slides to a list of image file paths.
|
|
///
|
|
/// Attempts LibreOffice headless conversion first. Falls back to
|
|
/// generating placeholder slide images (colored rectangles with slide numbers).
|
|
Future<List<String>> convertToImages(String pptxPath) async {
|
|
// Try LibreOffice first
|
|
final loImages = await _convertViaLibreOffice(pptxPath);
|
|
if (loImages.isNotEmpty) return loImages;
|
|
|
|
// Fallback: generate placeholder images
|
|
return _generatePlaceholderImages(pptxPath);
|
|
}
|
|
|
|
/// Open a file picker dialog and return the selected PPTX path, or null.
|
|
///
|
|
/// Uses the cross-platform file_picker package.
|
|
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;
|
|
}
|
|
|
|
// ---------------------------------------------------------------------------
|
|
// Implementation helpers
|
|
// ---------------------------------------------------------------------------
|
|
|
|
/// Extract text from PPTX slide XML by finding `<a:t>` content.
|
|
String _extractTextFromXml(String xml) {
|
|
final lines = <String>[];
|
|
// Match <a:t>...</a:t> — handles both <a:t>text</a:t> and <a:t xml:space="preserve">text</a:t>
|
|
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;
|
|
}
|
|
|
|
/// Try converting via LibreOffice headless.
|
|
Future<List<String>> _convertViaLibreOffice(String pptxPath) async {
|
|
try {
|
|
// Check if LibreOffice is available
|
|
final which = await Process.run('which', ['libreoffice']);
|
|
if (which.exitCode != 0) return [];
|
|
|
|
final outDir = await _makeTmpDir('pptx_images');
|
|
|
|
final result = await Process.run('libreoffice', [
|
|
'--headless',
|
|
'--convert-to',
|
|
'png',
|
|
'--outdir',
|
|
outDir.path,
|
|
pptxPath,
|
|
]);
|
|
|
|
if (result.exitCode != 0) return [];
|
|
|
|
// Collect generated PNGs, sorted by name
|
|
final pngs = await outDir
|
|
.list()
|
|
.where((f) => f.path.endsWith('.png'))
|
|
.map((f) => f.path)
|
|
.toList();
|
|
|
|
pngs.sort();
|
|
|
|
// Move to a persistent temp location so outDir can be cleaned up
|
|
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_${i + 1}.png');
|
|
await src.copy(dst);
|
|
persistentPaths.add(dst);
|
|
}
|
|
|
|
// Clean up the LibreOffice output dir
|
|
try {
|
|
await outDir.delete(recursive: true);
|
|
} catch (_) {}
|
|
|
|
return persistentPaths;
|
|
} catch (_) {
|
|
return [];
|
|
}
|
|
}
|
|
|
|
/// Generate placeholder slide images when LibreOffice is not available.
|
|
///
|
|
/// Uses ImageMagick `convert` to create PNG files with slide numbers.
|
|
/// If ImageMagick is not available, writes minimal 1x1 white PNGs as
|
|
/// last-resort placeholders.
|
|
Future<List<String>> _generatePlaceholderImages(String pptxPath) async {
|
|
// Count slides by unzipping and counting slide XML files
|
|
final slideCount = await _countSlides(pptxPath);
|
|
if (slideCount == 0) return [];
|
|
|
|
final outDir = await _makeTmpDir('pptx_placeholders');
|
|
final paths = <String>[];
|
|
|
|
// Try ImageMagick
|
|
final hasConvert = await _hasCommand('convert');
|
|
|
|
for (var i = 1; i <= slideCount; i++) {
|
|
final path = p.join(outDir.path, 'slide_$i.png');
|
|
if (hasConvert) {
|
|
await _generateWithImageMagick(path, i, slideCount);
|
|
} else {
|
|
await _writeMinimalPng(path);
|
|
}
|
|
paths.add(path);
|
|
}
|
|
|
|
return paths;
|
|
}
|
|
|
|
Future<int> _countSlides(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$')))
|
|
.length;
|
|
return count;
|
|
} catch (_) {
|
|
return 0;
|
|
} finally {
|
|
try {
|
|
await tmpDir.delete(recursive: true);
|
|
} catch (_) {}
|
|
}
|
|
}
|
|
|
|
Future<bool> _hasCommand(String cmd) async {
|
|
try {
|
|
final result = await Process.run('which', [cmd]);
|
|
return result.exitCode == 0;
|
|
} catch (_) {
|
|
return false;
|
|
}
|
|
}
|
|
|
|
Future<void> _generateWithImageMagick(
|
|
String outPath,
|
|
int slideNum,
|
|
int total,
|
|
) async {
|
|
// Light pastel background with slide number
|
|
final hue = ((slideNum - 1) * 137) % 360; // golden-angle spacing
|
|
await Process.run('convert', [
|
|
'-size',
|
|
'1920x1080',
|
|
'xc:hsl($hue, 60%, 92%)',
|
|
'-gravity',
|
|
'center',
|
|
'-pointsize',
|
|
'120',
|
|
'-fill',
|
|
'hsl($hue, 30%, 40%)',
|
|
'-annotate',
|
|
'+0+0',
|
|
'Slide $slideNum / $total',
|
|
outPath,
|
|
]);
|
|
}
|
|
|
|
/// Write a minimal valid 1x1 white PNG as an absolute last resort.
|
|
/// This is a hand-crafted PNG (IHDR + single white pixel IDAT + IEND).
|
|
Future<void> _writeMinimalPng(String path) async {
|
|
// Minimal valid 1x1 white PNG
|
|
const pngBytes = <int>[
|
|
0x89, 0x50, 0x4E, 0x47, 0x0D, 0x0A, 0x1A, 0x0A, // PNG signature
|
|
// IHDR chunk
|
|
0x00, 0x00, 0x00, 0x0D, // length = 13
|
|
0x49, 0x48, 0x44, 0x52, // "IHDR"
|
|
0x00, 0x00, 0x00, 0x01, // width = 1
|
|
0x00, 0x00, 0x00, 0x01, // height = 1
|
|
0x08, 0x02, // bit depth = 8, color type = 2 (RGB)
|
|
0x00, 0x00, 0x00, // compression, filter, interlace
|
|
0x90, 0x77, 0x53, 0xDE, // CRC
|
|
// IDAT chunk
|
|
0x00, 0x00, 0x00, 0x0C, // length = 12
|
|
0x49, 0x44, 0x41, 0x54, // "IDAT"
|
|
0x08, 0xD7, 0x63, 0xF8, 0xCF, 0xC0, 0x00, 0x00,
|
|
0x01, 0x01, 0x01, 0x00, // compressed data
|
|
0x18, 0xDD, 0x8D, 0xB4, // CRC
|
|
// IEND chunk
|
|
0x00, 0x00, 0x00, 0x00, // length = 0
|
|
0x49, 0x45, 0x4E, 0x44, // "IEND"
|
|
0xAE, 0x42, 0x60, 0x82, // CRC
|
|
];
|
|
await File(path).writeAsBytes(pngBytes);
|
|
}
|
|
|
|
Future<Directory> _makeTmpDir(String prefix) async {
|
|
final base = await getTemporaryDirectory();
|
|
final dir = Directory(p.join(base.path, '${prefix}_${_uuid.v4()}'));
|
|
await dir.create(recursive: true);
|
|
return dir;
|
|
}
|
|
}
|