Spaces:
Sleeping
Sleeping
File size: 2,011 Bytes
9f6318c f3612a8 b0828d2 f3612a8 44b22d1 fe5abc7 b1f06e2 69aaa24 1e57c9c f3612a8 b1f06e2 f3612a8 69aaa24 b0828d2 a2e0f7d b0828d2 56c42a3 b0828d2 69aaa24 73b8b37 f3612a8 73b8b37 b0828d2 a2e0f7d f3612a8 a2e0f7d b0828d2 56c42a3 b0828d2 a2e0f7d b0828d2 d7d509b b0828d2 56c42a3 b0828d2 d7d509b 9f6318c 73b8b37 f3612a8 9f6318c 810a8a7 b0828d2 1e57c9c b0828d2 f3612a8 1e57c9c | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 | import unicodedata
import logging
from .fixed_output import run_fixed_output
from .ai_output import run_ai_output
TRIGGER_MAP = {}
def set_trigger_map(mapping):
global TRIGGER_MAP
TRIGGER_MAP = mapping
logging.warning(f"[AGENT] TRIGGER_MAP loaded: {TRIGGER_MAP}")
def normalize(s):
return unicodedata.normalize("NFKC", s)
def clean(s):
s = unicodedata.normalize("NFKC", s)
return "".join(ch for ch in s if ch.isprintable())
def run_agent(user_input: str, patient_id: int, db):
logging.warning(f"[AGENT] user_input = {user_input}")
token = clean(user_input.split()[-1])
logging.warning(f"[AGENT] token = {token} (repr={repr(token)})")
# -----------------------------
# 1. prefix 補全（支援多選）
# -----------------------------
prefix_candidates = []
for skill_name, triggers in TRIGGER_MAP.items():
for t in triggers:
t_clean = clean(t)
if t_clean.startswith(token) and token != t_clean:
prefix_candidates.append(t_clean)
if len(prefix_candidates) > 1:
return {
"type": "trigger-multi-prefix",
"prefix": token,
"candidates": prefix_candidates
}
if len(prefix_candidates) == 1:
return {
"type": "trigger-prefix",
"prefix": token,
"full": prefix_candidates[0]
}
# -----------------------------
# 2. 完整 trigger
# -----------------------------
for skill_name, triggers in TRIGGER_MAP.items():
for t in triggers:
if clean(t) == token:
logging.warning(f"[AGENT] → full trigger match: {t}")
return run_fixed_output(skill_name)
# -----------------------------
# 3. fallback → AI（帶 patient_id + db）
# -----------------------------
logging.warning("[AGENT] → fallback to AI")
return run_ai_output(
input_text=user_input,
patient_id=patient_id,
db=db
) |