Improve learning discovery and dashboard i18n

app/actuators/lifecycle.py

@@ -46,6 +46,8 @@ _STOPWORDS = frozenset(
         "entity",
         "humidity",
         "illuminance",
+        "led",
+        "lidl",
         "light",
         "licht",
         "lichtschalter",
@@ -138,6 +140,33 @@ _AUTO_CONTEXT_CLASSES = frozenset({
     "presence",
     "window",
 })
+_PRESENCE_TOKENS = frozenset({
+    "besetzt",
+    "occupied",
+    "occupancy",
+    "presence",
+    "prasenz",
+    "praesenz",
+    "motion",
+    "bewegung",
+    "bewegungsmelder",
+})
+_MAILBOX_TOKENS = frozenset({"briefkasten", "mailbox", "post"})
+_CABINET_TOKENS = frozenset({"schrank", "cabinet"})
+_PV_TOKENS = frozenset({
+    "pv",
+    "solar",
+    "photovoltaik",
+    "akku",
+    "batterie",
+    "battery",
+    "einspeisung",
+    "wechselrichter",
+    "inverter",
+    "netzbezug",
+    "grid",
+    "verbrauch",
+})
 
 
 class ActuatorReconciliationService:
@@ -864,6 +893,18 @@ def _has_context_relationship(actuator: HaEntitySummary, entity: HaEntitySummary
         return True
     if _metadata_tokens(actuator).intersection(_metadata_tokens(entity)):
         return True
+    actuator_tokens = _metadata_tokens(actuator, include_stopwords=True)
+    entity_tokens = _metadata_tokens(entity, include_stopwords=True)
+    if _is_mailbox_reset_candidate(actuator_tokens, entity_tokens, entity):
+        return True
+    if actuator.domain in {"fan", "humidifier"} and (
+        _is_presence_context(entity) or entity.device_class in {"humidity", "moisture"}
+    ):
+        return True
+    if actuator.domain in {"climate", "cover", "fan", "humidifier", "light", "switch"} and (
+        entity_tokens.intersection(_PV_TOKENS)
+    ):
+        return True
     entity_tokens = _metadata_tokens(entity, include_stopwords=True)
     return bool(
         entity_tokens.intersection(_OUTDOOR_TOKENS)
@@ -878,6 +919,18 @@ def _eligible_for_auto_context(
     device_class = candidate.device_class or ""
     if device_class in _AUTO_CONTEXT_CLASSES:
         return True
+    if actuator.domain in {"fan", "humidifier"} and device_class in {
+        "humidity",
+        "moisture",
+        "temperature",
+    }:
+        return True
+    if actuator.domain in {"fan", "humidifier", "light", "switch"} and _is_presence_candidate(candidate):
+        return True
+    actuator_tokens = _metadata_tokens(actuator, include_stopwords=True)
+    candidate_tokens = _candidate_tokens(candidate, include_stopwords=True)
+    if _is_mailbox_reset_candidate(actuator_tokens, candidate_tokens, candidate):
+        return True
     if (
         actuator.device_name
         and candidate.device_name
@@ -899,6 +952,7 @@ def _score_candidate(
     score = 0.0
     actuator_tokens = _metadata_tokens(actuator)
     entity_tokens = _metadata_tokens(entity)
+    full_entity_tokens = _metadata_tokens(entity, include_stopwords=True)
     overlap = sorted(actuator_tokens.intersection(entity_tokens))
     if overlap:
         score += min(0.4, 0.1 * len(overlap))
@@ -924,6 +978,31 @@ def _score_candidate(
     if entity.device_class in preferred_device_classes:
         score += 0.2
         evidence.append(f"Passende device_class: {entity.device_class}")
+    if context and actuator.domain in {"fan", "humidifier"} and entity.device_class in {
+        "humidity",
+        "moisture",
+    }:
+        score += 0.3
+        evidence.append("Luftfeuchtigkeit ist primärer Kontext für Lüftung.")
+    if not context and actuator.domain in {"fan", "humidifier"} and entity.device_class in {
+        "humidity",
+        "moisture",
+    }:
+        score += 0.3
+        evidence.append("Luftfeuchtigkeit ist primärer Messwert für Lüftung.")
+    if context and actuator.domain in {"fan", "humidifier", "light", "switch"} and _is_presence_context(entity):
+        score += 0.3
+        evidence.append("Anwesenheit/Belegung ist primärer Schaltkontext.")
+    if context and _is_mailbox_reset_candidate(
+        _metadata_tokens(actuator, include_stopwords=True),
+        _metadata_tokens(entity, include_stopwords=True),
+        entity,
+    ):
+        score += 0.45
+        evidence.append("Briefkasten-Reset passt zur Schrank-/Entnahme-Tür.")
+    if full_entity_tokens.intersection(_PV_TOKENS):
+        score += 0.12 if context else 0.18
+        evidence.append("PV-/Akku-/Verbrauchswert ist als Energiemanagement-Kontext relevant.")
     if not context and actuator.domain == "light" and entity.device_class == "illuminance":
         score += 0.2
         evidence.append("Beleuchtungsstärke wird für Lichtaktoren bevorzugt.")
@@ -1068,11 +1147,83 @@ def _metadata_tokens(entity: HaEntitySummary, *, include_stopwords: bool = False
     for value in raw_values:
         if value is None:
             continue
-        for token in _TOKEN_PATTERN.findall(value.lower().replace("_", " ")):
-            if len(token) < 3 or (not include_stopwords and token in _STOPWORDS):
+        for token in _TOKEN_PATTERN.findall(_normalize_text(value)):
+            if (len(token) < 3 and token != "wc") or (not include_stopwords and token in _STOPWORDS):
                 continue
             tokens.add(token)
-    return tokens
+    return _expand_room_tokens(tokens)
+
+
+def _candidate_tokens(
+    candidate: AssignmentCandidate,
+    *,
+    include_stopwords: bool = False,
+) -> set[str]:
+    raw_values = [
+        candidate.entity_id,
+        candidate.friendly_name,
+        candidate.area_name,
+        candidate.device_name,
+    ]
+    tokens: set[str] = set()
+    for value in raw_values:
+        if value is None:
+            continue
+        for token in _TOKEN_PATTERN.findall(_normalize_text(value)):
+            if (len(token) < 3 and token != "wc") or (not include_stopwords and token in _STOPWORDS):
+                continue
+            tokens.add(token)
+    return _expand_room_tokens(tokens)
+
+
+def _expand_room_tokens(tokens: set[str]) -> set[str]:
+    expanded = set(tokens)
+    if "gaste" in expanded:
+        expanded.add("gaeste")
+    if {"gaste", "wc"}.issubset(expanded) or {"gaeste", "wc"}.issubset(expanded):
+        expanded.add("gaestewc")
+    if {"gaeste", "zimmer"}.issubset(expanded):
+        expanded.add("gaestezimmer")
+    return expanded
+
+
+def _normalize_text(value: str) -> str:
+    return (
+        value.lower()
+        .replace("_", " ")
+        .replace("ä", "ae")
+        .replace("ö", "oe")
+        .replace("ü", "ue")
+        .replace("ß", "ss")
+    )
+
+
+def _is_presence_context(entity: HaEntitySummary) -> bool:
+    if entity.device_class in {"motion", "occupancy", "presence"}:
+        return True
+    return bool(_metadata_tokens(entity, include_stopwords=True).intersection(_PRESENCE_TOKENS))
+
+
+def _is_presence_candidate(candidate: AssignmentCandidate) -> bool:
+    if candidate.device_class in {"motion", "occupancy", "presence"}:
+        return True
+    return bool(_candidate_tokens(candidate, include_stopwords=True).intersection(_PRESENCE_TOKENS))
+
+
+def _is_mailbox_reset_candidate(
+    actuator_tokens: set[str],
+    context_tokens: set[str],
+    entity: HaEntitySummary | AssignmentCandidate,
+) -> bool:
+    if not actuator_tokens.intersection(_MAILBOX_TOKENS):
+        return False
+    if not context_tokens.intersection(_CABINET_TOKENS):
+        return False
+    return entity.domain == "binary_sensor" and entity.device_class in {
+        "door",
+        "garage_door",
+        "opening",
+    }
 
 
 def _history_signature(sensor_id: str, points: list[NumericHistoryPoint]) -> str: