Raw PyTorch/Docling stack traces are no longer shown to the user.
Common failures (missing compiler, OOM, lock contention, corrupted
document) are mapped to actionable messages. Unknown errors are
truncated to 200 chars.
Ref #57 (M1)
TableFormerMode.ACCURATE is very expensive on CPU (~3-5x slower).
The default can now be set to "fast" on resource-constrained
environments (HF Spaces) via the DEFAULT_TABLE_MODE env var.
User-specified table_mode in the request still takes precedence.
Ref #57 (H1)
Prevents PyTorch/Docling pipeline crashes on HF Spaces CPU by:
- Reducing max file size from 50 MB to 5 MB
- Adding configurable MAX_PAGE_COUNT setting (env var, default unlimited)
- Increasing conversion timeout from 600s to 900s
- Adding frontend upload validation with explicit error messages
- Exposing maxPageCount via /api/health for dynamic UI hints
Domain model state transitions, repository CRUD operations, and service
methods now have descriptive docstrings to lower the contribution
barrier for an open-source project.
Read uploaded files in fixed-size chunks instead of a single file.read()
to reduce peak memory pressure. Also enforces the size limit during
streaming so oversized payloads are rejected before fully buffered.
Unbounded asyncio.create_task calls could exhaust CPU and memory on
modest hardware. Add a configurable semaphore (MAX_CONCURRENT_ANALYSES,
default 3) so excess jobs queue instead of running all at once.
Replace broad `except Exception` with specific types (FileNotFoundError,
PermissionError, OSError) so errors are properly categorized in logs.
Users reporting bugs will get actionable messages instead of generic ones.
UPLOAD_DIR and DB_PATH were read directly from os.environ, bypassing
the Settings dataclass. This caused an inconsistency where overriding
Settings had no effect on these values. Now all modules import from
infra.settings.settings.
Extract bounding boxes from chunk doc_items provenance in the chunker,
propagate through domain/service/API layers, and render highlighted
bboxes on canvas when hovering a chunk card. Reset highlights on
mode and page changes to prevent stale visual state.
AnalysisService gains rechunk() and inline chunking during conversion.
ChunkingOptionsRequest/ChunkResponse schemas, POST rechunk endpoint,
and conditional chunker injection in main.py (local engine only).
- Remove dead get_analysis_service() function in main.py
- Scope file deletion to UPLOAD_DIR to prevent path traversal
- Parse datetime strings back to datetime objects in repos
- Add 10-minute polling timeout in frontend analysis store
- Accept .pdf extension (not just MIME type) on drag-and-drop
- Guard localStorage access for private browsing compatibility
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Bug #1: _on_task_done now receives job_id via functools.partial and
calls _mark_failed when the background task raises or is cancelled,
preventing jobs from being stuck in RUNNING state forever.
Bug #5: _parse_response wraps json.loads in try/except JSONDecodeError
so malformed json_content strings fall back gracefully instead of crashing.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Extract domain value objects and ports from parsing.py, move Docling-specific
code to infra/local_converter.py, and convert analysis_service to a class
with injected DocumentConverter. This prepares the codebase for plugging in
alternative conversion backends (e.g. Docling Serve) via the Protocol pattern.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
