Fix: Revert to inline shared types (Turbopack can't resolve path aliases)
Some checks failed
CI / go-lint (push) Has been skipped
CI / python-lint (push) Has been skipped
CI / nodejs-lint (push) Has been skipped
CI / test-go-school (push) Successful in 40s
CI / test-go-edu-search (push) Successful in 30s
CI / test-python-klausur (push) Failing after 2m56s
CI / test-python-agent-core (push) Successful in 19s
CI / test-nodejs-website (push) Successful in 25s

Turbopack doesn't support tsconfig path aliases pointing outside
the project root. Reverted to copying shared types directly into
each service. The canonical source remains shared/types/*.ts,
synced via scripts/sync-shared-types.sh.

Changes:
- Reverted docker-compose.yml contexts to ./service
- Reverted Dockerfiles to simple COPY . .
- Removed @shared/* from tsconfigs
- Removed symlinks + .gitignore hacks
- Added scripts/sync-shared-types.sh for keeping copies in sync

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
Benjamin Admin
2026-04-25 17:13:18 +02:00
parent 2eb17fd349
commit b49ee3467e
24 changed files with 3127 additions and 47 deletions

View File

@@ -1 +1,127 @@
export * from '@shared/types/ocr-labeling'
/**
* Shared TypeScript types for OCR Labeling UI.
*
* Single source of truth used by:
* - admin-lehrer (ai/ocr-labeling)
* - website (admin/ocr-labeling)
*/
/**
* Available OCR Models
*
* - llama3.2-vision:11b: Vision LLM, beste Qualitaet bei Handschrift (Standard)
* - trocr: Microsoft TrOCR, schnell bei gedrucktem Text
* - paddleocr: PaddleOCR + LLM, 4x schneller durch Hybrid-Ansatz
* - donut: Document Understanding Transformer, strukturierte Dokumente
*/
export type OCRModel = 'llama3.2-vision:11b' | 'trocr' | 'paddleocr' | 'donut'
export const OCR_MODEL_INFO: Record<OCRModel, { label: string; description: string; speed: string }> = {
'llama3.2-vision:11b': {
label: 'Vision LLM',
description: 'Beste Qualitaet bei Handschrift',
speed: 'langsam',
},
trocr: {
label: 'Microsoft TrOCR',
description: 'Schnell bei gedrucktem Text',
speed: 'schnell',
},
paddleocr: {
label: 'PaddleOCR + LLM',
description: 'Hybrid-Ansatz: OCR + Strukturierung',
speed: 'sehr schnell',
},
donut: {
label: 'Donut',
description: 'Document Understanding fuer Tabellen/Formulare',
speed: 'mittel',
},
}
export interface OCRSession {
id: string
name: string
source_type: 'klausur' | 'handwriting_sample' | 'scan'
description?: string
ocr_model?: OCRModel
total_items: number
labeled_items: number
confirmed_items: number
corrected_items: number
skipped_items: number
created_at: string
}
export interface OCRItem {
id: string
session_id: string
session_name: string
image_path: string
image_url?: string
ocr_text?: string
ocr_confidence?: number
ground_truth?: string
status: 'pending' | 'confirmed' | 'corrected' | 'skipped'
metadata?: Record<string, unknown>
created_at: string
}
export interface OCRStats {
total_sessions?: number
session_id?: string
name?: string
total_items: number
labeled_items: number
confirmed_items: number
corrected_items: number
skipped_items?: number
pending_items: number
exportable_items?: number
accuracy_rate: number
avg_label_time_seconds?: number
progress_percent?: number
}
export interface TrainingSample {
id: string
image_path: string
ground_truth: string
export_format: 'generic' | 'trocr' | 'llama_vision'
training_batch: string
exported_at?: string
}
export interface CreateSessionRequest {
name: string
source_type: 'klausur' | 'handwriting_sample' | 'scan'
description?: string
ocr_model?: OCRModel
}
export interface ConfirmRequest {
item_id: string
label_time_seconds?: number
}
export interface CorrectRequest {
item_id: string
ground_truth: string
label_time_seconds?: number
}
export interface ExportRequest {
export_format: 'generic' | 'trocr' | 'llama_vision'
session_id?: string
batch_id?: string
}
export interface UploadResult {
id: string
filename: string
image_path: string
image_hash: string
ocr_text?: string
ocr_confidence?: number
status: string
}