Spaces:
Sleeping
Sleeping
Rivalcoder
commited on
Commit
·
785fe24
1
Parent(s):
e0753c4
Add application file
Browse files- app.py +20 -0
- requirements.txt +3 -0
app.py
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from fastapi import FastAPI, File, UploadFile
|
| 2 |
+
import fitz # PyMuPDF
|
| 3 |
+
import uvicorn
|
| 4 |
+
|
| 5 |
+
app = FastAPI()
|
| 6 |
+
|
| 7 |
+
@app.post("/extract-text")
|
| 8 |
+
async def extract_text(file: UploadFile = File(...)):
|
| 9 |
+
try:
|
| 10 |
+
contents = await file.read()
|
| 11 |
+
doc = fitz.open(stream=contents, filetype="pdf")
|
| 12 |
+
|
| 13 |
+
extracted_text = ""
|
| 14 |
+
for i, page in enumerate(doc):
|
| 15 |
+
extracted_text += f"\n\n--- Page {i + 1} ---\n\n" + page.get_text()
|
| 16 |
+
|
| 17 |
+
return {"filename": file.filename, "text": extracted_text}
|
| 18 |
+
|
| 19 |
+
except Exception as e:
|
| 20 |
+
return {"error": str(e)}
|
requirements.txt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
fastapi
|
| 2 |
+
uvicorn
|
| 3 |
+
PyMuPDF
|