Upload schemas/letter.json with huggingface_hub

schemas/letter.json

@@ -0,0 +1,73 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "$id": "german-ocr-3/schemas/letter.json",
+  "title": "GermanOCR3 Letter (Brief)",
+  "description": "Schema fuer deutsche Geschaefts- oder Behoerdenbriefe (DIN-5008-naher Aufbau).",
+  "type": "object",
+  "additionalProperties": false,
+  "required": ["document_type", "language", "subject", "body"],
+  "properties": {
+    "document_type": {"const": "letter"},
+    "language": {"type": "string", "default": "de"},
+
+    "sender": {
+      "type": ["object", "null"],
+      "additionalProperties": false,
+      "properties": {
+        "name": {"type": ["string", "null"]},
+        "address": {"type": ["string", "null"]},
+        "email": {"type": ["string", "null"]},
+        "phone": {"type": ["string", "null"]}
+      }
+    },
+    "recipient": {
+      "type": ["object", "null"],
+      "additionalProperties": false,
+      "properties": {
+        "name": {"type": ["string", "null"]},
+        "address": {"type": ["string", "null"]}
+      }
+    },
+
+    "place": {"type": ["string", "null"]},
+    "date": {"description": "Briefdatum YYYY-MM-DD", "type": ["string", "null"]},
+    "reference": {
+      "description": "Aktenzeichen / Ihr Zeichen / Unser Zeichen.",
+      "type": ["object", "null"],
+      "additionalProperties": false,
+      "properties": {
+        "ihr_zeichen": {"type": ["string", "null"]},
+        "ihre_nachricht_vom": {"type": ["string", "null"]},
+        "unser_zeichen": {"type": ["string", "null"]},
+        "unsere_nachricht_vom": {"type": ["string", "null"]}
+      }
+    },
+
+    "subject": {"type": ["string", "null"]},
+    "salutation": {"type": ["string", "null"]},
+    "body": {
+      "description": "Kompletter Brieftext, Absaetze mit \\n\\n getrennt.",
+      "type": ["string", "null"]
+    },
+    "closing": {"type": ["string", "null"]},
+    "signatory": {"type": ["string", "null"]},
+    "enclosures": {"type": "array", "items": {"type": "string"}, "default": []},
+
+    "deadlines_mentioned": {
+      "type": "array",
+      "default": [],
+      "items": {
+        "type": "object",
+        "additionalProperties": false,
+        "properties": {
+          "date": {"type": ["string", "null"]},
+          "context": {"type": ["string", "null"]}
+        }
+      }
+    },
+
+    "raw_text": {"type": ["string", "null"]},
+    "confidence": {"type": ["number", "null"], "minimum": 0, "maximum": 1},
+    "notes": {"type": "array", "items": {"type": "string"}, "default": []}
+  }
+}