Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
|
@@ -106,19 +106,23 @@ def read_pdf_online(url):
|
|
| 106 |
print(f"reading {url}")
|
| 107 |
response = requests.get(url, stream=True)
|
| 108 |
text=""
|
|
|
|
| 109 |
if response.status_code == 200:
|
| 110 |
-
|
| 111 |
-
f.
|
| 112 |
-
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
|
| 116 |
-
|
| 117 |
-
|
| 118 |
-
|
| 119 |
-
|
| 120 |
-
|
| 121 |
-
|
|
|
|
|
|
|
|
|
|
| 122 |
else:
|
| 123 |
text = response.status_code
|
| 124 |
print(text)
|
|
|
|
| 106 |
print(f"reading {url}")
|
| 107 |
response = requests.get(url, stream=True)
|
| 108 |
text=""
|
| 109 |
+
|
| 110 |
if response.status_code == 200:
|
| 111 |
+
try:
|
| 112 |
+
with open(f"{uid}.pdf", "wb") as f:
|
| 113 |
+
f.write(response.content)
|
| 114 |
+
#f.close()
|
| 115 |
+
#out = Path("./data.pdf")
|
| 116 |
+
#print (out)
|
| 117 |
+
reader = PdfReader(f"{uid}.pdf")
|
| 118 |
+
number_of_pages = len(reader.pages)
|
| 119 |
+
for i in range(number_of_pages-1):
|
| 120 |
+
page = reader.pages[i]
|
| 121 |
+
text = f'{text}\n{page.extract_text()}'
|
| 122 |
+
print(f"PDF_TEXT:: {text}")
|
| 123 |
+
return text
|
| 124 |
+
except Exception as e:
|
| 125 |
+
return e
|
| 126 |
else:
|
| 127 |
text = response.status_code
|
| 128 |
print(text)
|