Spaces:
Paused
Paused
Update pdftotext.py
Browse files- pdftotext.py +0 -2
pdftotext.py
CHANGED
|
@@ -144,8 +144,6 @@ def apiFiltering(apitext):
|
|
| 144 |
def clean_text(text):
|
| 145 |
# Replace all newlines and tabs with a space
|
| 146 |
text = re.sub(r'[\n\t]+', ' ', text)
|
| 147 |
-
# Collapse multiple spaces into one
|
| 148 |
-
text = re.sub(r'\s+', ' ', text)
|
| 149 |
return text.strip()
|
| 150 |
|
| 151 |
def texts_from_pdfAllText(link):
|
|
|
|
| 144 |
def clean_text(text):
|
| 145 |
# Replace all newlines and tabs with a space
|
| 146 |
text = re.sub(r'[\n\t]+', ' ', text)
|
|
|
|
|
|
|
| 147 |
return text.strip()
|
| 148 |
|
| 149 |
def texts_from_pdfAllText(link):
|