python：wordpress系统百度ai敏感词违规词检测替换-源码网

最近，遇到这个问题，怎么检查已经发布的文章里面的敏感词，用百度ai文本检测功能，有25万的免费额度，在网上找了很多插件，感觉都不好用，其中要能检测和替换敏感词的功能，于是，想到用python写个源码，直接链接wordpress数据库来处理，流程：

第一，链接wp数据库，定位数据表，提取数据

# Connect to the Empire CMS database
MYSQL_URL = '103.91.210.249'  # 数据库地址
MYSQL_PORT = 3306  # 数据库端口
MYSQL_USER = "qmwu_com"  # 数据库用户名
MYSQL_PASSWORD = "a7sX2DcPbWbanPWw"  # 数据库密码
MYSQL_DATABASE = "qmwu_com"  # 数据库名称
MYSQL_CHARSET = "utf8"  # 数据库字符编码
conn = pymysql.connect(host=MYSQL_URL, port=MYSQL_PORT, user=MYSQL_USER, password=MYSQL_PASSWORD,
                       database=MYSQL_DATABASE, charset=MYSQL_CHARSET)

定位数据表

sql = cursor.execute("SELECT id,post_content FROM qmwu_posts WHERE ischeck = 0  order by rand() limit 1 ")
if sql == 0:
    print("检测完毕！")
    break
articles = cursor.fetchone()
articles = list(articles)
ids = articles[0]  # 文章id
content = articles[1]  # 文章内容
if content == "":
    print(articles[0], "文章内容为空！跳过")
    cursor.execute("UPDATE qmwu_posts SET ischeck = 1 WHERE id=%s  ", ids)
    continue
spam = baidu_ai_sensitive_word_detection(content)

百度ai处理

url = "https://aip.baidubce.com/rest/2.0/solution/v1/text_censor/v2/user_defined?access_token=" + get_access_token()
tx = {"text": text}
tx = urlencode(tx)
payload = tx
headers = {
    'Content-Type': 'application/x-www-form-urlencoded',
    'Accept': 'application/json'
}

response = requests.request(“POST”, url, headers=headers, data=payload)
print(response.text)
result = response.json()
try:
if result[“conclusionType”] == 2: # 不合格
s = result[“data”]
s1 = s[0][‘msg’] # 违规类型
s2 = s[0][‘hits’][0][‘wordHitPositions’][0][‘keyword’] # 违规词
return s1, s2
else:
return False

spam = baidu_ai_sensitive_word_detection(content)
if not spam:
    print(f"ID:{ids} 的内容合规")
    cursor.execute("UPDATE qmwu_posts SET ischeck = 1 WHERE id=%s  ", ids)
else:
    print(f"ID:{ids} 的内容不合规")
    keyword = spam[1]
    content = content.replace(keyword, '***')
    cursor.execute("UPDATE qmwu_posts SET post_content = %s WHERE id = %s and ischeck = 0", (content, ids))
    print(spam[0] + ":替换【" + spam[1] + "】，为***")
    cursor.execute("UPDATE qmwu_posts SET ischeck = 1 WHERE id=%s  ", ids)
conn.commit()
# time.sleep(1)  # Wait 10 minutes before checking the database again