Commit
·
2c39311
1
Parent(s):
a420b55
Update app.py
Browse files
app.py
CHANGED
@@ -113,7 +113,7 @@ def util_get_list_page_and_passage(list_docs, list_save_path):
|
|
113 |
sections = split_into_sentences_with_offsets(page_document)
|
114 |
for sub_sub_index, sub_sub_item in enumerate(sections):
|
115 |
sub_text=sub_sub_item[0]
|
116 |
-
passage_document = Document(page_content=sub_text, metadata={"page_index": txt_index, "
|
117 |
passage_documents.append(passage_document)
|
118 |
return(page_documents,passage_documents)
|
119 |
|
@@ -150,7 +150,7 @@ def util_get_list_pageno_and_contents(some_query_passage, page_documents,passage
|
|
150 |
|
151 |
tmp_array = []
|
152 |
for i, x in enumerate(passage_nodes):
|
153 |
-
tmp_dict = {"passage_content":x.page_content, "page_no":x.metadata['page_index']+1, "page_content": passage_documents[x.metadata['page_index']], "score": rescore[i] }
|
154 |
df = pd.DataFrame(tmp_array)
|
155 |
|
156 |
|
|
|
113 |
sections = split_into_sentences_with_offsets(page_document)
|
114 |
for sub_sub_index, sub_sub_item in enumerate(sections):
|
115 |
sub_text=sub_sub_item[0]
|
116 |
+
passage_document = Document(page_content=sub_text, metadata={"page_index": txt_index, "file_name" : str(list_save_path[ind_doc])})
|
117 |
passage_documents.append(passage_document)
|
118 |
return(page_documents,passage_documents)
|
119 |
|
|
|
150 |
|
151 |
tmp_array = []
|
152 |
for i, x in enumerate(passage_nodes):
|
153 |
+
tmp_dict = {"passage_content":x.page_content, "page_no":x.metadata['page_index']+1, "file_name": x.metadata['file_name'], "page_content": passage_documents[x.metadata['page_index']], "score": rescore[i] }
|
154 |
df = pd.DataFrame(tmp_array)
|
155 |
|
156 |
|