From 22f370322412074174cde20ecfd14ec03657ab63 Mon Sep 17 00:00:00 2001 From: lyg <1543117173@qq.com> Date: 星期一, 07 七月 2025 16:20:25 +0800 Subject: [PATCH] 生成数据库 --- main.py | 34 ++++++++++++++++++++++++++++++++++ 1 files changed, 34 insertions(+), 0 deletions(-) diff --git a/main.py b/main.py index e69de29..6cc6b78 100644 --- a/main.py +++ b/main.py @@ -0,0 +1,34 @@ +import asyncio +import os +import sys + +from db_struct_flow import DbStructFlow, tc_data_generate +from knowledgebase.db.doc_db_helper import doc_dbh +from knowledgebase.doc.doc_processor import DocProcessor +from knowledgebase.doc.entity_helper import init_entity_helper + + +def doc_split(project_path): + docs_path = f'{project_path}/docs' + files = os.listdir(docs_path) + files = [f'{docs_path}/{x}' for x in filter(lambda x: x.endswith('.docx'), files)] + for file in files: + DocProcessor(file).process() + + + +def main(): + project_path = sys.argv[1] + if not project_path: + print("missing project path. eg: python main.py <path/to/project>") + return + # 鎷嗗垎鏂囨。 + doc_dbh.set_project_path(project_path) + init_entity_helper() + # doc_split(project_path) + # 鍚姩澶фā鍨嬪鐞嗘祦绋� + asyncio.run(DbStructFlow(project_path).run()) + # 鐢熸垚鎸囦护鏁版嵁琛� + tc_data_generate() +if __name__ == "__main__": + main() \ No newline at end of file -- Gitblit v1.9.1