- Raise quality acceptance threshold from 65→85 (configurable via TEXT_QUALITY_THRESHOLD)
- Reject text with significant issues (excessive_typos, garbage_characters,
incoherent_text, fragmented_sentences) even when score is above threshold
(configurable via TEXT_QUALITY_SIGNIFICANT_ISSUES)
- Add compare_text_quality() for AI-powered head-to-head comparison of
original embedded text vs fresh OCR output
- Update process_document to pass original text to OCR task for comparison
- Update process_with_ocr to run comparison and keep the higher-quality text
- Add new settings to settings_service.py metadata
- Update docs/ConfigurationGuide.md with new settings
- Add comprehensive tests for new threshold and comparison logic
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
- Update _compute_processing_flow to recognize process_with_ocr as the OCR
stage and remap legacy process_with_azure_document_intelligence log entries
for backward compatibility
- Normalize legacy OCR step name in _compute_step_summary log fallback
- Add process_with_ocr to REAL_MAIN_STEPS/REAL_STEPS in step_manager,
file_status, and file_queries (keeping legacy name for old DB entries)
- Update retry logic in api/files.py to retry failed OCR via process_with_ocr
(handles both step names as aliases)
- Fix process_document.py to log process_with_ocr as skipped (not azure step)
for the local text extraction path
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
Root cause: ensure_ocr_languages_from_settings() only downloaded tessdata
when the 'tesseract' provider was active, but embed_text_layer() uses
ocrmypdf (which needs tessdata) as a fallback for ALL OCR providers.
- embed_text_layer(): call ensure_tesseract_languages(language) after
confirming ocrmypdf is on PATH, so language data is present before
ocrmypdf is invoked (prevents exit code 3 for fra/deu/etc.)
- ensure_ocr_languages_from_settings(): extend the condition from
'tesseract' in active_providers to also trigger when ocrmypdf is
on PATH, enabling proactive pre-download at startup for any config
- Tests: mock shutil.which and ensure_tesseract_languages in affected
test cases; rename azure-only test and add new test for ocrmypdf case
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
- Add app/utils/ocr_language_manager.py: detects tessdata dir, downloads
missing .traineddata files via wget/curl from tessdata_fast GitHub repo,
pre-downloads EasyOCR models, exposes async background-thread helper
- TesseractOCRProvider.process() calls ensure_tesseract_languages() before
running pytesseract; raises clear error if languages remain unavailable
- EasyOCRProvider.process() logs informational message when models download
- app/main.py: calls ensure_ocr_languages_async() at startup
- app/utils/settings_sync.py: triggers language re-check after every
settings reload so UI changes take effect without container restart
- app/api/settings.py: adds POST /api/settings/install-ocr-languages
endpoint for on-demand language installation from the admin UI
- Dockerfile: adds wget for runtime tessdata downloads
- docs/ConfigurationGuide.md: documents automatic language download
- tests/test_ocr_language_manager.py: 29 unit tests
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
- PDFs are now uploaded to Mistral Files API (POST /v1/files) and
processed via a signed document_url, resolving the 422 error caused
by passing data:application/pdf;base64,... to an image endpoint
- Images (JPEG/PNG/GIF/WEBP/BMP/TIFF) use base64 image_url directly
- Unsupported MIME types raise a clear ValueError
- Magic-byte fallback detects PDFs with no file extension
- Switches from openai chat completions to requests HTTP calls
- Adds helper method _upload_pdf_and_get_document()
- Adds TestMistralOCRProvider with 9 unit tests covering all paths
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
A) Per-option Save Button
- Add per-setting Save button in settings.html (visible only when value changed)
- Button calls POST /api/settings/{key} directly; existing bulk Save retained
- Add Audit Log link in settings page header
B) Immediate Worker Sync
- New app/utils/settings_sync.py with notify_settings_updated() (Redis version key)
and register_settings_reload_signal() (Celery task_prerun handler)
- Register signal in celery_worker.py at startup
- All API write paths call notify_settings_updated() after successful saves
C) Audit Log
- Add SettingsAuditLog model (key, old_value, new_value, changed_by, changed_at, action)
- save_setting_to_db / delete_setting_from_db accept changed_by and write audit entries
- New get_audit_log() service function (masks sensitive values)
- New GET /api/settings/audit-log endpoint (admin-only)
- New GET /admin/settings/audit-log view + audit_log.html template
- Visible to all admins (per clarified requirement)
D) Config Rollback / History
- New get_setting_history() and rollback_setting() service functions
- New GET /api/settings/{key}/history endpoint
- New POST /api/settings/{key}/rollback/{history_id} endpoint
- Rollback buttons in audit_log.html with confirmation dialog
- Tests: 25 new tests covering audit log, rollback, worker sync helpers, and API endpoints
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
`GET /api/settings/{key}` was calling `validate_setting_key()` which raises
HTTP 404 for keys not in SETTING_METADATA. The test expects 200 with value=None
for unknown keys.
Added `validate_setting_key_format()` to `input_validation.py` that validates
only the key format without the SETTING_METADATA existence check. Updated
`get_setting` to use the format-only validator; POST/DELETE endpoints continue
using the full `validate_setting_key()` for write-side security.
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
- Add [[tool.mypy.overrides]] section for app/utils/** with disallow_untyped_defs=true
- Add type annotations to all functions in app/utils/ (12 files)
- Fix type annotations in app/config.py and app/database.py (imported by utils)
- All 85 source files now pass mypy type checking
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
- Add comprehensive tests for imap_tasks.py covering:
- Lock acquire/release mechanisms
- pull_all_inboxes task with various scenarios
- pull_inbox for Gmail and non-Gmail with edge cases
- find_all_mail_xlist functionality
- Extended fetch_attachments tests for all MIME types
- Edge cases: invalid JSON, missing Message-ID, already processed, etc.
- Achieve 98.26% coverage for imap_tasks.py (up from 48.78%)
- Fix config_validator.py to include validate_auth_config export
- Update tests to verify all exports including validate_auth_config
- Note: config_validator.py file is shadowed by config_validator/ directory
in Python's module resolution, so it cannot be directly imported or tested.
The package's __init__.py (which has 100% coverage) is what's actually used.
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
- Fix PLW2901: Use different variable name for stripped lines in loop
- Fix E721: Use 'is' instead of '==' for type comparisons
- Add noqa comments for intentional security warnings (S321, S507, S110, S603)
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
- Implement step timeout detection to prevent files from getting stuck in 'pending' state
- Add monitor_stalled_steps periodic task running every minute (Celery Beat)
- Automatically mark in-progress steps as failed if they exceed timeout (default: 10 minutes)
- Add step_timeout configuration setting (default: 600 seconds)
- Recover stalled steps with error message indicating when timeout was triggered
- Fix duplicate check to exclude self-comparison (file not duplicate of itself)
When processing crashes or hangs:
1. Worker detects stalled steps (in_progress for >10 minutes)
2. Marks them as failed with timeout error message
3. Updates UI to show failure status
4. Allows file to be retried or handled by user
This prevents files from being indefinitely stuck in processing state and provides
visibility into what went wrong.
- Add enable_deduplication and show_deduplication_step config options
- Rename hash_file step to check_for_duplicates
- Make deduplication step conditional based on configuration
- Add is_duplicate and duplicate_of_id fields to FileRecord model
- Create database migration for new deduplication fields
- Update process_document task to log deduplication results
- Update step visualization to show/hide step based on config
- Update status calculations to include deduplication step conditionally
- Default: deduplication enabled, step displayed
- Can be configured to hide from UI while still processing
- Add original_file_path and processed_file_path columns to FileRecord model
- Create database migration for new fields
- Implement get_unique_filepath_with_counter() with -0001 suffix format
- Update process_document to save immutable copy to /workdir/original
- Add force_cloud_ocr parameter to process_document for forced OCR
- Update embed_metadata to use new collision handling
- Update metadata JSON to include file path references
- Add /files/{file_id}/reprocess-with-cloud-ocr API endpoint
- Update processed_file_path in database during embedding
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
- Fix _compute_status_from_logs to track latest status per unique step
- Fix _compute_step_summary to count only latest status per step
- Add comprehensive tests for both fixes
- Resolves issue where completed files showed as "Processing"
- Resolves issue where metrics showed incorrect counts
Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
Christian Krakau-Louis