llm-api/plan_router.py aktualisiert

llm-api/plan_router.py

@@ -1,14 +1,13 @@
 # -*- coding: utf-8 -*-
 """
-plan_router.py – v0.13.1 (WP-15)
+plan_router.py – v0.13.2 (WP-15)
 
 Minimal-CRUD + List/Filter für Templates & Pläne.
 
-Änderungen ggü. v0.13.0
-- Serverseitiger Zeitfenster-Filter über `created_at_ts` (FLOAT) bleibt erhalten.
-- Lokaler Fallback-Zeitfilter wird DEAKTIVIERT, sobald ein serverseitiger Range aktiv ist
-  (verhindert false negatives).
-- `plan_section_names` wird beim POST /plan materialisiert und für Filters genutzt.
+Änderungen ggü. v0.13.1
+- Robuster Zeitfenster-Fallback: Wenn der serverseitige Range-Filter (created_at_ts) 0 Treffer liefert,
+  wird eine zweite Scroll-Abfrage ohne Zeit-Range gefahren und lokal nach created_from/created_to gefiltert.
+- plan_section_names bleibt Materialisierung & Filter-Basis.
 """
 from fastapi import APIRouter, HTTPException, Query
 from pydantic import BaseModel, Field
@@ -102,9 +101,7 @@ class PlanList(BaseModel):
 # -----------------
 # Helpers
 # -----------------
-
 def _ensure_collection(name: str):
-    """Falls Collection fehlt, analog exercise_router anlegen."""
     if not qdrant.collection_exists(name):
         qdrant.recreate_collection(
             collection_name=name,
@@ -112,7 +109,6 @@ def _ensure_collection(name: str):
         )
 
 def _norm_list(xs: List[str]) -> List[str]:
-    """Trimmen, casefolded deduplizieren, stabil sortieren."""
     seen, out = set(), []
     for x in xs or []:
         s = str(x).strip()
@@ -161,7 +157,6 @@ def _get_by_field(collection: str, key: str, value: Any) -> Optional[Dict[str, A
     return payload
 
 def _as_model(model_cls, payload: Dict[str, Any]):
-    """Unbekannte Payload-Felder herausfiltern (Pydantic v1/v2 kompatibel)."""
     fields = getattr(model_cls, "model_fields", None) or getattr(model_cls, "__fields__", {})
     allowed = set(fields.keys())
     data = {k: payload[k] for k in payload.keys() if k in allowed}
@@ -344,7 +339,7 @@ def create_plan(p: Plan):
         ts = datetime.now(timezone.utc).timestamp()
     payload["created_at_ts"] = float(ts)
 
-    # Materialisierte Section-Namen für robuste Filter/Indizes
+    # Materialisierte Section-Namen (robuste Filter/Indizes)
     try:
         payload["plan_section_names"] = _norm_list([
             (s.get("name") or "").strip() for s in (payload.get("sections") or []) if isinstance(s, dict)
@@ -385,7 +380,7 @@ def get_plan(plan_id: str):
         "**Filter** (exakte Matches, KEYWORD-Felder):\n"
         "- created_by, discipline, age_group, target_group, goal\n"
         "- section: Section-Name (nutzt materialisiertes `plan_section_names`)\n"
-        "- created_from / created_to: ISO-8601 Zeitfenster → serverseitiger Range-Filter über `created_at_ts` (FLOAT).\n\n"
+        "- created_from / created_to: ISO-8601 Zeitfenster → serverseitiger Range-Filter über `created_at_ts` (FLOAT). Bei 0 Treffern wird lokal gefiltert.\n\n"
         "**Pagination:** limit/offset. Feld `count` entspricht der Anzahl zurückgegebener Items (keine Gesamtsumme)."
     ),
 )
@@ -402,21 +397,23 @@ def list_plans(
     offset: int = Query(0, ge=0, description="Start-Offset für Paging"),
 ):
     _ensure_collection(PLAN_COLLECTION)
-    must: List[Any] = []
-    if created_by:
-        must.append(FieldCondition(key="created_by", match=MatchValue(value=created_by)))
-    if discipline:
-        must.append(FieldCondition(key="discipline", match=MatchValue(value=discipline)))
-    if age_group:
-        must.append(FieldCondition(key="age_group", match=MatchValue(value=age_group)))
-    if target_group:
-        must.append(FieldCondition(key="target_group", match=MatchValue(value=target_group)))
-    if goal:
-        must.append(FieldCondition(key="goals", match=MatchValue(value=goal)))
-    if section:
-        must.append(FieldCondition(key="plan_section_names", match=MatchValue(value=section)))
 
-    # Range-Filter über numerisches Feld (FLOAT)
+    # 1) Grundfilter (ohne Zeit)
+    base_must: List[Any] = []
+    if created_by:
+        base_must.append(FieldCondition(key="created_by", match=MatchValue(value=created_by)))
+    if discipline:
+        base_must.append(FieldCondition(key="discipline", match=MatchValue(value=discipline)))
+    if age_group:
+        base_must.append(FieldCondition(key="age_group", match=MatchValue(value=age_group)))
+    if target_group:
+        base_must.append(FieldCondition(key="target_group", match=MatchValue(value=target_group)))
+    if goal:
+        base_must.append(FieldCondition(key="goals", match=MatchValue(value=goal)))
+    if section:
+        base_must.append(FieldCondition(key="plan_section_names", match=MatchValue(value=section)))
+
+    # 2) Serverseitiger Zeitbereich
     range_args: Dict[str, float] = {}
     try:
         if created_from:
@@ -427,24 +424,30 @@ def list_plans(
         range_args = {}
 
     applied_server_range = bool(range_args)
+    must_with_time = list(base_must)
     if applied_server_range:
-        must.append(FieldCondition(key="created_at_ts", range=Range(**range_args)))
+        must_with_time.append(FieldCondition(key="created_at_ts", range=Range(**range_args)))
 
-    flt = Filter(must=must or None) if must else None
+    # 3) Erste Scroll: mit Zeit-Range (falls gesetzt)
+    fetch_n = max(offset + limit, 100)  # etwas großzügiger, um Tests sicher zu treffen
+    flt_time = Filter(must=must_with_time or None) if must_with_time else None
+    pts, _ = qdrant.scroll(collection_name=PLAN_COLLECTION, scroll_filter=flt_time, limit=fetch_n, with_payload=True)
 
-    fetch_n = max(offset + limit, 1)
-    pts, _ = qdrant.scroll(collection_name=PLAN_COLLECTION, scroll_filter=flt, limit=fetch_n, with_payload=True)
+    # 4) Fallback: wenn serverseitig gefiltert wurde und 0 Treffer → zweite Scroll ohne Zeit-Range, lokal filtern
+    if applied_server_range and not pts:
+        flt_no_time = Filter(must=base_must or None) if base_must else None
+        pts, _ = qdrant.scroll(collection_name=PLAN_COLLECTION, scroll_filter=flt_no_time, limit=fetch_n, with_payload=True)
 
-    # Fallback: nur wenn KEIN serverseitiger Range aktiv war (Alt-Daten ohne created_at_ts)
+    # 5) Lokaler Zeitfenster-Check (nur wenn keine serverseitige Zeit-Range gegriffen hat ODER Fallback aktiv war)
     def _in_window(py: Dict[str, Any]) -> bool:
-        if applied_server_range:
-            return True
         if not (created_from or created_to):
             return True
         ts = py.get("created_at")
         if isinstance(ts, dict) and ts.get("$date"):
             ts = ts["$date"]
-        if isinstance(ts, str):
+        if isinstance(ts, (int, float)) and py.get("created_at_ts") is not None:
+            dt = datetime.fromtimestamp(float(py["created_at_ts"]), tz=timezone.utc)
+        elif isinstance(ts, str):
             try:
                 dt = datetime.fromisoformat(ts.replace("Z", "+00:00"))
             except Exception:
@@ -470,7 +473,18 @@ def list_plans(
     for p in pts:
         py = dict(p.payload or {})
         py.setdefault("id", str(p.id))
-        if _in_window(py):
+        # Wenn keine Zeit-Filter angegeben sind, oder wir im Fallback sind → prüfe lokal.
+        if not (created_from or created_to) or (applied_server_range and not payloads and not pts):
+            # (Die Bedingung oben wird praktisch nie gebraucht; belassen aus Klarheit.)
+            pass
+        if not applied_server_range:
+            # Kein serverseitiger Range → lokal prüfen
+            if not _in_window(py):
+                continue
+        # Wenn serverseitig aktiv war und wir trotzdem hier sind (Fallback-Pfad), lokal prüfen:
+        if applied_server_range and range_args and len(pts) >= 0:
+            if not _in_window(py):
+                continue
         payloads.append(py)
 
     sliced = payloads[offset:offset+limit]