File tree Expand file tree Collapse file tree 1 file changed +5
-2
lines changed
backend/src/document_sources Expand file tree Collapse file tree 1 file changed +5
-2
lines changed Original file line number Diff line number Diff line change @@ -43,13 +43,17 @@ def get_documents_from_file_by_path(file_path,file_name):
43
43
44
44
if page .metadata ['page_number' ]> page_number :
45
45
page_number += 1
46
+ if not metadata :
47
+ metadata = {'total_pages' :unstructured_pages [- 1 ].metadata ['page_number' ]}
46
48
pages .append (Document (page_content = page_content , metadata = metadata ))
47
49
page_content = ''
48
50
49
51
if page == unstructured_pages [- 1 ]:
52
+ if not metadata :
53
+ metadata = {'total_pages' :unstructured_pages [- 1 ].metadata ['page_number' ]}
50
54
pages .append (Document (page_content = page_content , metadata = metadata ))
51
55
52
- elif page .metadata ['category' ]== 'PageBreak' :
56
+ elif page .metadata ['category' ]== 'PageBreak' and page != unstructured_pages [ 0 ] :
53
57
page_number += 1
54
58
pages .append (Document (page_content = page_content , metadata = metadata ))
55
59
page_content = ''
@@ -65,5 +69,4 @@ def get_documents_from_file_by_path(file_path,file_name):
65
69
else :
66
70
logging .info (f'File { file_name } does not exist' )
67
71
raise Exception (f'File { file_name } does not exist' )
68
-
69
72
return file_name , pages , file_extension
You can’t perform that action at this time.
0 commit comments