Spaces:
Runtime error
Runtime error
Joe Hare
commited on
Commit
•
ba05b5a
1
Parent(s):
4d331dd
add ml-based version of this
Browse files- app.py +42 -0
- requirements.txt +3 -0
app.py
ADDED
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
import pandas as pd
|
3 |
+
from transformers import pipeline
|
4 |
+
import base64
|
5 |
+
|
6 |
+
def upload_and_extract_text():
|
7 |
+
uploaded_file = st.file_uploader("Upload PDF file", type=["pdf"])
|
8 |
+
|
9 |
+
if uploaded_file is not None:
|
10 |
+
ocr_pipeline = pipeline("text2text-generation", model="google/t5-v1_1-large")
|
11 |
+
extracted_text = ocr_pipeline(uploaded_file.read(), max_length=1024, do_sample=False)[0]["generated_text"]
|
12 |
+
return extracted_text
|
13 |
+
|
14 |
+
def text_to_dataframe(text):
|
15 |
+
lines = text.split("\n")
|
16 |
+
data = []
|
17 |
+
for line in lines:
|
18 |
+
data.append([line])
|
19 |
+
df = pd.DataFrame(data, columns=["Text"])
|
20 |
+
return df
|
21 |
+
|
22 |
+
def main():
|
23 |
+
st.title("PDF to Spreadsheet Converter")
|
24 |
+
st.write("Upload a PDF file to extract text and save it as a spreadsheet.")
|
25 |
+
|
26 |
+
extracted_text = upload_and_extract_text()
|
27 |
+
|
28 |
+
if extracted_text is not None:
|
29 |
+
st.write("### Extracted Text")
|
30 |
+
st.write(extracted_text)
|
31 |
+
|
32 |
+
df = text_to_dataframe(extracted_text)
|
33 |
+
st.write("### Spreadsheet Preview")
|
34 |
+
st.write(df)
|
35 |
+
|
36 |
+
csv = df.to_csv(index=False)
|
37 |
+
b64 = base64.b64encode(csv.encode()).decode()
|
38 |
+
href = f'<a href="data:file/csv;base64,{b64}" download="extracted_text.csv">Download Extracted Text as CSV</a>'
|
39 |
+
st.markdown(href, unsafe_allow_html=True)
|
40 |
+
|
41 |
+
if __name__ == "__main__":
|
42 |
+
main()
|
requirements.txt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
streamlit==1.3.0
|
2 |
+
transformers==4.13.0
|
3 |
+
pandas==1.4.1
|