marcel/familienarchiv
1
0
Fork 0
Code Issues 125 Pull Requests Actions Packages Projects Releases Wiki Activity

docs(c4): accuracy audit — split L3 diagrams, add 6 new sub-diagrams, fix all stale content #448

Merged
marcel merged 27 commits from docs/post-refactor-accuracy-audit into main 2026-05-06 20:00:16 +02:00
Conversation 47 Commits 27 Files Changed 1 +6 -1
1 changed files with 6 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit 35d82814a6 - Show all commits

7
docs/architecture/c4-diagrams.md
View File

@@ -277,6 +277,9 @@ C4Component
Component(ocrTraining, "OcrTrainingService", "Spring Service", "Orchestrates model training: exports training data as ZIP, calls Python /train or /segtrain, persists training metrics in OcrTrainingRunRepository.")
}
Component(transcriptionSvc, "TranscriptionService", "Spring Service", "See diagram 3b.2. Called by OcrAsyncRunner to persist transcription blocks per page.")
Component(annotationSvc, "AnnotationService", "Spring Service", "See diagram 3b.2. Called by OcrAsyncRunner to persist OCR-generated annotation regions per page.")
Rel(frontend, ocrCtrl, "OCR trigger, status, and progress requests", "HTTP / JSON / SSE")
Rel(ocrCtrl, ocrSvc, "Single-document jobs", "")
Rel(ocrCtrl, ocrBatch, "Batch jobs", "")
@@ -286,7 +289,9 @@ C4Component
Rel(ocrAsync, ocrClient, "Streams OCR results page by page", "HTTP / NDJSON")
Rel(ocrTraining, ocrClient, "Sends training data ZIP", "HTTP / multipart")
Rel(ocrClient, ocrPy, "POST /ocr/stream, /train, /segtrain, /train-sender", "HTTP / REST")
Rel(ocrAsync, db, "Reads / writes job state, transcription blocks, annotations", "JDBC")
Rel(ocrAsync, transcriptionSvc, "Saves transcription blocks per page", "")
Rel(ocrAsync, annotationSvc, "Saves annotation regions per page", "")
Rel(ocrAsync, db, "Reads / writes OCR job state", "JDBC")
Rel(ocrAsync, minio, "Generates presigned URLs for PDF fetch", "S3 API")
Rel(ocrPy, minio, "Fetches PDF via presigned URL", "HTTP / S3 presigned")
Rel(ocrTraining, db, "Persists training run metrics", "JDBC")