feat: clean pdf fitz text
This commit is contained in:
@@ -1,5 +1,5 @@
|
||||
from predict import predict_no_ui
|
||||
from toolbox import CatchException, report_execption, write_results_to_file, predict_no_ui_but_counting_down
|
||||
from toolbox import CatchException, report_execption, write_results_to_file, predict_no_ui_but_counting_down, clean_text
|
||||
fast_debug = False
|
||||
|
||||
|
||||
@@ -11,6 +11,7 @@ def 解析PDF(file_manifest, project_folder, top_p, temperature, chatbot, histor
|
||||
file_content = ""
|
||||
for page in doc:
|
||||
file_content += page.get_text()
|
||||
file_content = clean_text(file_content)
|
||||
print(file_content)
|
||||
|
||||
prefix = "接下来请你逐文件分析下面的论文文件,概括其内容" if index==0 else ""
|
||||
|
||||
Reference in New Issue
Block a user