Merge remote-tracking branch 'origin/Django' into Django
This commit is contained in:
@@ -17,8 +17,10 @@ class ElasticConfig(AppConfig):
|
||||
return
|
||||
|
||||
# 延迟导入，避免循环导入或过早加载
|
||||
from .es_connect import create_index_with_mapping
|
||||
from .es_connect import create_index_with_mapping, get_type_list
|
||||
try:
|
||||
create_index_with_mapping()
|
||||
types = get_type_list()
|
||||
print(f"🔎 启动时 type_list: {types}")
|
||||
except Exception as e:
|
||||
print(f"❌ ES 初始化失败: {e}")
|
||||
@@ -6,6 +6,8 @@ ACHIEVEMENT_INDEX = Index(ACHIEVEMENT_INDEX_NAME)
|
||||
ACHIEVEMENT_INDEX.settings(number_of_shards=1, number_of_replicas=0)
|
||||
USER_INDEX = Index(USER_INDEX_NAME)
|
||||
USER_INDEX.settings(number_of_shards=1, number_of_replicas=0)
|
||||
GLOBAL_INDEX = Index(GLOBAL_INDEX_NAME)
|
||||
GLOBAL_INDEX.settings(number_of_shards=1, number_of_replicas=0)
|
||||
|
||||
|
||||
|
||||
@@ -37,3 +39,10 @@ class UserDocument(Document):
|
||||
model = User
|
||||
# fields列表应该只包含需要特殊处理的字段，或者可以完全省略
|
||||
# 因为我们已经显式定义了所有字段
|
||||
|
||||
@GLOBAL_INDEX.doc_type
|
||||
class GlobalDocument(Document):
|
||||
type_list = fields.KeywordField()
|
||||
|
||||
class Django:
|
||||
model = ElasticNews
|
||||
|
||||
@@ -4,8 +4,8 @@ Django版本的ES连接和操作模块
|
||||
"""
|
||||
from elasticsearch import Elasticsearch
|
||||
from elasticsearch_dsl import connections
|
||||
from .documents import AchievementDocument, UserDocument
|
||||
from .indexes import ACHIEVEMENT_INDEX_NAME, USER_INDEX_NAME
|
||||
from .documents import AchievementDocument, UserDocument, GlobalDocument
|
||||
from .indexes import ACHIEVEMENT_INDEX_NAME, USER_INDEX_NAME, GLOBAL_INDEX_NAME
|
||||
import hashlib
|
||||
import time
|
||||
|
||||
@@ -17,6 +17,7 @@ es = connections.get_connection()
|
||||
|
||||
DATA_INDEX_NAME = ACHIEVEMENT_INDEX_NAME
|
||||
USERS_INDEX_NAME = USER_INDEX_NAME
|
||||
GLOBAL_TYPES_INDEX_NAME = GLOBAL_INDEX_NAME
|
||||
|
||||
def create_index_with_mapping():
|
||||
"""创建索引和映射配置（仅当索引不存在时）"""
|
||||
@@ -36,7 +37,25 @@ def create_index_with_mapping():
|
||||
else:
|
||||
print(f"ℹ️ 索引 {USERS_INDEX_NAME} 已存在，跳过创建")
|
||||
|
||||
# --- 3. 创建默认管理员用户（可选：也可检查用户是否已存在）---
|
||||
# --- 3. 处理全局类型索引 ---
|
||||
if not es.indices.exists(index=GLOBAL_TYPES_INDEX_NAME):
|
||||
GlobalDocument.init()
|
||||
default_types = ['软著', '专利', '奖状']
|
||||
doc = GlobalDocument(type_list=default_types)
|
||||
doc.meta.id = 'types'
|
||||
doc.save()
|
||||
print(f"✅ 创建索引 {GLOBAL_TYPES_INDEX_NAME} 并写入默认类型")
|
||||
else:
|
||||
try:
|
||||
GlobalDocument.get(id='types')
|
||||
except Exception:
|
||||
default_types = ['软著', '专利', '奖状']
|
||||
doc = GlobalDocument(type_list=default_types)
|
||||
doc.meta.id = 'types'
|
||||
doc.save()
|
||||
print("ℹ️ 全局类型文档缺失，已补充默认类型")
|
||||
|
||||
# --- 4. 创建默认管理员用户（可选：也可检查用户是否已存在）---
|
||||
# 这里简单处理：每次初始化都写入（可能重复），建议加唯一性判断
|
||||
admin_user = {
|
||||
"user_id": 0,
|
||||
@@ -57,6 +76,36 @@ def create_index_with_mapping():
|
||||
print(f"❌ 创建索引失败: {str(e)}")
|
||||
# raise # 可选：在 AppConfig 中捕获，这里可以 re-raise 便于调试
|
||||
|
||||
def get_type_list():
|
||||
try:
|
||||
doc = GlobalDocument.get(id='types')
|
||||
lst = [str(t).strip().strip(';') for t in (doc.type_list or [])]
|
||||
return lst
|
||||
except Exception:
|
||||
return ['软著', '专利', '奖状']
|
||||
|
||||
def ensure_type_in_list(type_name: str):
|
||||
if not type_name:
|
||||
return False
|
||||
norm = str(type_name).strip().strip(';')
|
||||
try:
|
||||
try:
|
||||
doc = GlobalDocument.get(id='types')
|
||||
cur = list(doc.type_list or [])
|
||||
except Exception:
|
||||
cur = ['软著', '专利', '奖状']
|
||||
doc = GlobalDocument(type_list=cur)
|
||||
doc.meta.id = 'types'
|
||||
cur_sanitized = {str(t).strip().strip(';') for t in cur}
|
||||
if norm not in cur_sanitized:
|
||||
cur.append(norm)
|
||||
doc.type_list = cur
|
||||
doc.save()
|
||||
return True
|
||||
return False
|
||||
except Exception:
|
||||
return False
|
||||
|
||||
def get_doc_id(data):
|
||||
"""
|
||||
根据数据内容生成唯一ID（用于去重）
|
||||
|
||||
@@ -1,4 +1,5 @@
|
||||
INDEX_NAME = "wordsearch266666"
|
||||
USER_NAME = "users"
|
||||
INDEX_NAME = "wordsearch266666789"
|
||||
USER_NAME = "users_123"
|
||||
ACHIEVEMENT_INDEX_NAME = INDEX_NAME
|
||||
USER_INDEX_NAME = USER_NAME
|
||||
GLOBAL_INDEX_NAME = "global11111"
|
||||
|
||||
@@ -244,13 +244,14 @@ def ocr_and_extract_info(image_path: str):
|
||||
base_url="https://aistudio.baidu.com/llm/lmapi/v3"
|
||||
client = OpenAI(api_key=api_key, base_url=base_url)
|
||||
|
||||
types = get_type_list()
|
||||
chat_completion = client.chat.completions.create(
|
||||
messages=[
|
||||
{"role": "system", "content": "你是一个能理解图片和文本的助手，请根据用户提供的信息进行回答。"},
|
||||
{
|
||||
"role": "user",
|
||||
"content": [
|
||||
{"type": "text", "text": "请识别这张图片中的信息，将你认为重要的数据转换为不包含嵌套的json，不要显示其它信息以便于解析直接输出json结果即可你可以自行决定使用哪些json字段"},
|
||||
{"type": "text", "text": f"请识别这张图片中的信息，将你认为重要的数据转换为不包含嵌套的json，不要显示其它信息以便于解析，直接输出json结果即可。使用“数据类型”字段表示这个东西的大致类型，除此之外你可以自行决定使用哪些json字段。“数据类型”的内容有严格规定，请查看{json.dumps(types, ensure_ascii=False)}中是否包含你所需要的类型，确定不包含后你才可以填入你觉得合适的大致分类。"},
|
||||
{"type": "image_url", "image_url": {"url": f"data:image/png;base64,{base64_image}"}},
|
||||
],
|
||||
},
|
||||
@@ -370,6 +371,7 @@ def confirm(request):
|
||||
if not isinstance(edited, dict) or not edited:
|
||||
return JsonResponse({"status": "error", "message": "数据不能为空"}, status=400)
|
||||
|
||||
ensure_type_in_list(edited.get("数据类型"))
|
||||
to_store = {
|
||||
"writer_id": str(request.session.get("user_id")),
|
||||
"data": json_to_string(edited),
|
||||
|
||||
Reference in New Issue
Block a user