174 lines
8.2 KiB
Python
174 lines
8.2 KiB
Python
#!/usr/bin/env python3
|
||
"""Update ability labels for speaking_qa records based on question content."""
|
||
import json, requests, time
|
||
|
||
CRED = "/root/.openclaw/credentials/xiaoyan/config.json"
|
||
TOKEN_APP = "CMHSbUUjka3TrUsaxxEc297ongf"
|
||
TBL = "tblRGv7k4WH58Jgq"
|
||
|
||
L_INFO = "基础信息表达｜个人信息问答"
|
||
L_PREF = "表达喜好与理由"
|
||
L_PAST = "过去经历描述｜Past Activities"
|
||
L_INTR = "互动应答｜问答交流"
|
||
L_EXCH = "信息交换｜双向问答"
|
||
|
||
def classify(question):
|
||
q = question.lower()
|
||
|
||
# 互动应答
|
||
if any(kw in q for kw in ["what do you say", "what will you say", "what would you say",
|
||
"what shouldn't we do", "what must we not do", "what should we do",
|
||
"what do we do next", "what do the shop workers say"]):
|
||
return L_INTR
|
||
|
||
# 过去经历描述
|
||
if any(kw in q for kw in ["what did you", "when did you last", "who did you help",
|
||
"last weekend", "what have you learned", "what have you decided",
|
||
"what toy did you buy", "where did you meet",
|
||
"what did your father say", "what did you first find",
|
||
"what gift have you got", "what did you see",
|
||
"what you're glad you could do"]):
|
||
return L_PAST
|
||
|
||
# 信息交换
|
||
if any(kw in q for kw in ["what do we need", "what things are there",
|
||
"what will your friends bring", "what do people use",
|
||
"what do we use", "what shall we", "whose bag",
|
||
"whose schoolbag", "whose pen is this", "who is this gift for"]):
|
||
return L_EXCH
|
||
|
||
# 个人信息: name, age, from, family, appearance
|
||
if any(kw in q for kw in ["what's your name", "how old", "where are you from",
|
||
"who do you live", "is your home in", "what rooms",
|
||
"what does your dad look like", "what does your mother look like",
|
||
"what is your mom's hair", "what are your mom's",
|
||
"what does your sister wear", "what does a dinosaur look like",
|
||
"what do the clouds look like", "what is the moon like",
|
||
"what's your best friend like", "what does your cat look like",
|
||
"how is the bear", "who is your best", "who is taller",
|
||
"what are you good at", "what is your best friend good at",
|
||
"what is your main study goal", "what is your goal",
|
||
"what habit is very important", "what information",
|
||
"what do you have to do every day",
|
||
"tell me what you're glad", "please introduce your pet",
|
||
"what pet do you love", "any animals. which one",
|
||
"what toys do you have", "how many books does your dad have",
|
||
"what is in your schoolbag", "what's in your schoolbag",
|
||
"what can you find", "what is there in your schoolbag",
|
||
"what is in your classroom", "what are in the classroom"]):
|
||
return L_INFO
|
||
|
||
# 信息交换 fallback
|
||
if any(kw in q for kw in ["where is your book", "is this pen yours", "where are your books",
|
||
"which page", "where should you put", "can you swim",
|
||
"is there a book", "can you help me", "what do you say to make",
|
||
"what do you say when others", "what do you say when your friend",
|
||
"what do you say when you want", "what do you say to draw",
|
||
"what do you say to introduce"]):
|
||
return L_INTR
|
||
|
||
# 表达喜好
|
||
if any(kw in q for kw in ["favorite", "favourite", "what do you like", "what kind of",
|
||
"what do you think", "what do you want", "do you like",
|
||
"what do you enjoy", "what books do you", "do you keep",
|
||
"are your classmates", "are you afraid",
|
||
"what do your family plan", "what do your parents plan",
|
||
"what do you plan", "what are you going to",
|
||
"what will you", "what would you", "what can you do",
|
||
"how do you like", "what colour do you like", "what colour do you want",
|
||
"what gift do you want", "what do you want to eat", "why do you like",
|
||
"which car is faster", "what sport can't you do",
|
||
"what can you see", "who will you sit with",
|
||
"how is the ice cream", "what animal might it be",
|
||
"what is the rabbit eating", "what is the bird",
|
||
"what do you need", "what will we have at the party",
|
||
"what must we protect", "what do you believe",
|
||
"what makes your dad unhappy", "what do your friends",
|
||
"whose pen do you think", "what hair has your dad",
|
||
"what sports can you do", "what do you need for",
|
||
"if you can't find", "where will we go", "what do people use",
|
||
"the radio has no sound", "what do you need to do"]):
|
||
return L_PREF
|
||
|
||
return L_INFO
|
||
|
||
# Main
|
||
with open(CRED) as f:
|
||
cred = json.load(f)
|
||
aid = cred["apps"][0]["appId"]
|
||
asecret = cred["apps"][0]["appSecret"]
|
||
|
||
def get_token():
|
||
r = requests.post("https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal",
|
||
json={"app_id": aid, "app_secret": asecret})
|
||
return r.json()["tenant_access_token"]
|
||
|
||
token = get_token()
|
||
|
||
all_items = []
|
||
page_token = None
|
||
while True:
|
||
url = f"https://open.feishu.cn/open-apis/bitable/v1/apps/{TOKEN_APP}/tables/{TBL}/records?page_size=50"
|
||
if page_token:
|
||
url += f"&page_token={page_token}"
|
||
resp = requests.get(url, headers={"Authorization": f"Bearer {token}"})
|
||
data = resp.json()
|
||
all_items.extend(data["data"].get("items", []))
|
||
if not data["data"].get("has_more"):
|
||
break
|
||
page_token = data["data"].get("page_token")
|
||
|
||
updates = 0
|
||
stats = {}
|
||
for item in all_items:
|
||
rid = item["record_id"]
|
||
r = requests.get(
|
||
f"https://open.feishu.cn/open-apis/bitable/v1/apps/{TOKEN_APP}/tables/{TBL}/records/{rid}",
|
||
headers={"Authorization": f"Bearer {token}"})
|
||
fields = r.json()["data"]["record"]["fields"]
|
||
fid = fields.get("题目集合 ID", "")
|
||
jd = fields.get("jsonData", "")
|
||
|
||
if not jd or fid in ("L1", "L2", None):
|
||
continue
|
||
|
||
try:
|
||
j = json.loads(jd)
|
||
except:
|
||
continue
|
||
|
||
changed = False
|
||
for grp_name in ["first", "second"]:
|
||
if grp_name not in j:
|
||
continue
|
||
for q in j[grp_name].get("questionSet", []):
|
||
existing = q.get("ability", [])
|
||
valid_labels = {L_INFO, L_PREF, L_PAST, L_INTR, L_EXCH}
|
||
if existing and all(a in valid_labels for a in existing):
|
||
continue
|
||
|
||
content = q.get("content", q.get("question", ""))
|
||
label = classify(content)
|
||
q["ability"] = [label]
|
||
changed = True
|
||
stats[label] = stats.get(label, 0) + 1
|
||
|
||
if changed:
|
||
new_jd = json.dumps(j, ensure_ascii=False)
|
||
r = requests.put(
|
||
f"https://open.feishu.cn/open-apis/bitable/v1/apps/{TOKEN_APP}/tables/{TBL}/records/{rid}",
|
||
headers={"Authorization": f"Bearer {token}", "Content-Type": "application/json"},
|
||
json={"fields": {"jsonData": new_jd}})
|
||
code = r.json().get("code")
|
||
if code == 0:
|
||
updates += 1
|
||
print(f"✅ {fid}: updated")
|
||
else:
|
||
print(f"❌ {fid}: code={code}")
|
||
time.sleep(0.3)
|
||
|
||
print(f"\nUpdated {updates} records")
|
||
print("\nLabel distribution:")
|
||
for label, count in sorted(stats.items(), key=lambda x: -x[1]):
|
||
print(f" {label}: {count}")
|