feat: белый список форматов + сводка батча по типам

deploy/services/classify.py

@@ -103,10 +103,11 @@ def classify_batch(batch_id, llm_client=None, repo=None):
     garbage = 0
     json_fixes = 0
     json_total = 0
+    type_counts = {}  # doc_type → count for batch summary
 
     def _classify_one(doc):
         """Классифицировать один документ: фильтр → выжимка → LLM → сохранить."""
-        nonlocal garbage, json_fixes, json_total
+        nonlocal garbage, json_fixes, json_total, type_counts
         try:
             # Stage 1: garbage by filename (0 tokens)
             if _is_garbage_by_filename(doc["filename"]):
@@ -127,6 +128,8 @@ def classify_batch(batch_id, llm_client=None, repo=None):
             json_total += 1
             if needed_fix:
                 json_fixes += 1
+            dtype = result.get("doc_type", "other")
+            type_counts[dtype] = type_counts.get(dtype, 0) + 1
             _db.set_classification(
                 doc["id"],
                 result.get("doc_type", "other"),
@@ -151,7 +154,8 @@ def classify_batch(batch_id, llm_client=None, repo=None):
                 failed += 1
 
     return {"ok": True, "total": total, "done": done, "failed": failed,
-            "garbage": garbage, "json_fix_rate": round(json_fixes / max(json_total, 1), 3)}
+            "garbage": garbage, "json_fix_rate": round(json_fixes / max(json_total, 1), 3),
+            "types": type_counts, "summary": f"{total} total, {done} classified, {failed} failed, {garbage} garbage"}
 
 
 def _safe_json_parse(raw):

deploy/services/upload.py

@@ -4,6 +4,15 @@ from db import documents, contracts, supplements
 from .parse import parse_file
 
 MAX_FILE_SIZE = 100 * 1024 * 1024  # 100 MB
+ALLOWED_EXTENSIONS = {"pdf", "docx", "doc", "zip"}
+
+
+def _check_format(filename: str) -> str | None:
+    """Validate file extension. Returns error string or None."""
+    ext = filename.rsplit(".", 1)[-1].lower() if "." in filename else ""
+    if ext not in ALLOWED_EXTENSIONS:
+        return f"unsupported format: .{ext} (allowed: {', '.join(sorted(ALLOWED_EXTENSIONS))})"
+    return None
 
 
 # ── Чистые функции (тестируемы без HTTP) ──────────────────────
@@ -77,6 +86,10 @@ def store_document(filename: str, file_data: bytes, contract_id: str = "",
     if not filename or not file_data:
         return {"ok": False, "error": "no file"}
 
+    fmt_err = _check_format(filename)
+    if fmt_err:
+        return {"ok": False, "error": fmt_err}
+
     import hashlib
     content_hash = hashlib.sha256(file_data).hexdigest()