Create data/pdf_reader.py
Browse files- data/pdf_reader.py +31 -0
data/pdf_reader.py
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# pdf_reader.py
|
| 2 |
+
import PyPDF2
|
| 3 |
+
from typing import List
|
| 4 |
+
|
| 5 |
+
class PDFReader:
|
| 6 |
+
def __init__(self):
|
| 7 |
+
self.page_list = []
|
| 8 |
+
|
| 9 |
+
def read_pdf(self, file_path: str) -> List[str]:
|
| 10 |
+
"""
|
| 11 |
+
Read PDF content and return list of pages
|
| 12 |
+
Each element in the list is the text content of a page
|
| 13 |
+
"""
|
| 14 |
+
try:
|
| 15 |
+
# Open and read the PDF file
|
| 16 |
+
with open(file_path, 'rb') as file:
|
| 17 |
+
pdf_reader = PyPDF2.PdfReader(file)
|
| 18 |
+
num_pages = len(pdf_reader.pages)
|
| 19 |
+
|
| 20 |
+
# Extract text from each page
|
| 21 |
+
self.page_list = []
|
| 22 |
+
for page_num in range(num_pages):
|
| 23 |
+
page = pdf_reader.pages[page_num]
|
| 24 |
+
text = page.extract_text()
|
| 25 |
+
if text: # Only add non-empty pages
|
| 26 |
+
self.page_list.append(text.strip())
|
| 27 |
+
|
| 28 |
+
return self.page_list
|
| 29 |
+
|
| 30 |
+
except Exception as e:
|
| 31 |
+
raise Exception(f"Error reading PDF: {str(e)}")
|