Spaces:
Sleeping
Sleeping
Update pdfparser_performance.py
Browse files- pdfparser_performance.py +6 -6
pdfparser_performance.py
CHANGED
@@ -14,11 +14,11 @@ def rename_and_move_files(root_folder):
|
|
14 |
def process_pdf_performance(path):
|
15 |
path_extracted_pdf = path+"/extracted_pdf/"
|
16 |
os.makedirs(path_extracted_pdf, exist_ok=True)
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
rename_and_move_files(path_extracted_pdf)
|
24 |
|
|
|
14 |
def process_pdf_performance(path):
|
15 |
path_extracted_pdf = path+"/extracted_pdf/"
|
16 |
os.makedirs(path_extracted_pdf, exist_ok=True)
|
17 |
+
for root, dirs, files in os.walk(path):
|
18 |
+
for file in files:
|
19 |
+
if file.endswith('.pdf'):
|
20 |
+
print("FILE IS ", os.path.join(root, file))
|
21 |
+
file_proper_format = file.replace(" ", "\ ")
|
22 |
+
os.system(f"marker_single {os.path.join(root, file_proper_format)} {path_extracted_pdf} --batch_multiplier 10")
|
23 |
rename_and_move_files(path_extracted_pdf)
|
24 |
|