Files
Akiba So 72428dc075
Some checks failed
CI / Test (Server, optional) (push) Failing after 2m10s
Windows Build / Build Windows (x64) (push) Failing after 29s
CI / Test (Flutter, Linux) (push) Has been cancelled
CI / Analyze (Flutter) (push) Has been cancelled
Fix bugs across app + server, optimize UI/UX, add Gitea CI
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>
2026-06-21 03:18:00 +08:00

79 lines
2.6 KiB
Python

"""OCR engine for BadNote using EasyOCR.
Lightweight handwriting-capable OCR using EasyOCR with CPU-only inference.
Suitable for Zen2 APU 25W / 16GB RAM (~200MB memory once loaded).
"""
import asyncio
import logging
logger = logging.getLogger(__name__)
# NOTE: `easyocr` (and its torch dependency) is heavy and optional. It is
# imported lazily inside the engine so that importing the FastAPI app — and
# running its test suite — does not require the OCR dependencies. Install them
# with `pip install -r requirements-ocr.txt` when running the OCR worker.
class OcrEngine:
"""EasyOCR-based text recognition engine.
Lazy-loads the reader on first use to avoid startup overhead.
Supports Chinese (simplified) + English. Runs on CPU only.
"""
def __init__(self):
self._reader = None
def _ensure_reader(self):
"""Lazy-initialize the EasyOCR reader."""
if self._reader is None:
import easyocr # imported lazily; see module docstring note
logger.info("Loading EasyOCR reader (ch_sim + en, CPU)...")
self._reader = easyocr.Reader(['ch_sim', 'en'], gpu=False)
logger.info("EasyOCR reader loaded")
async def recognize(self, image_bytes: bytes) -> str:
"""Recognize text from image bytes.
Args:
image_bytes: Raw image file bytes (PNG, JPEG, etc.)
Returns:
Recognized text as a single string, or empty string on failure.
"""
if not image_bytes:
return ""
loop = asyncio.get_event_loop()
try:
self._ensure_reader()
results = await loop.run_in_executor(
None, self._reader.readtext, image_bytes
)
# results is a list of (bbox, text, confidence) tuples
text_parts = [text for _, text, _ in results if text.strip()]
return ' '.join(text_parts)
except Exception as exc:
logger.error("OCR recognition failed: %s", exc)
return ""
async def recognize_file(self, image_path: str) -> str:
"""Recognize text from an image file path.
Args:
image_path: Path to the image file on disk.
Returns:
Recognized text as a single string, or empty string on failure.
"""
try:
with open(image_path, 'rb') as f:
image_bytes = f.read()
return await self.recognize(image_bytes)
except Exception as exc:
logger.error("Failed to read image file %s: %s", image_path, exc)
return ""