Spaces:
Runtime error
Runtime error
Upload 2 files
Browse files- app.py +21 -0
- requirements.txt +67 -0
app.py
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import fitz # PyMuPDF
|
| 2 |
+
import pytesseract
|
| 3 |
+
import gradio as gr
|
| 4 |
+
from PIL import Image
|
| 5 |
+
|
| 6 |
+
def pdf_to_text(pdf_file):
|
| 7 |
+
doc = fitz.open(pdf_file)
|
| 8 |
+
text = ""
|
| 9 |
+
for page in doc:
|
| 10 |
+
pix = page.get_pixmap()
|
| 11 |
+
img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
|
| 12 |
+
text += pytesseract.image_to_string(img)
|
| 13 |
+
doc.close()
|
| 14 |
+
return text
|
| 15 |
+
|
| 16 |
+
def pdf_to_text_interface(pdf_file):
|
| 17 |
+
text = pdf_to_text(pdf_file)
|
| 18 |
+
return text
|
| 19 |
+
|
| 20 |
+
iface = gr.Interface(fn=pdf_to_text_interface, inputs="file", outputs="text", title="PDF to Text Converter")
|
| 21 |
+
iface.launch()
|
requirements.txt
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
aiofiles==23.1.0
|
| 2 |
+
aiohttp==3.8.4
|
| 3 |
+
aiosignal==1.3.1
|
| 4 |
+
altair==5.0.1
|
| 5 |
+
anyio==3.7.1
|
| 6 |
+
async-timeout==4.0.2
|
| 7 |
+
attrs==23.1.0
|
| 8 |
+
certifi==2023.5.7
|
| 9 |
+
charset-normalizer==3.1.0
|
| 10 |
+
click==8.1.3
|
| 11 |
+
colorama==0.4.6
|
| 12 |
+
contourpy==1.1.0
|
| 13 |
+
cycler==0.11.0
|
| 14 |
+
Cython==0.29.36
|
| 15 |
+
fastapi==0.99.1
|
| 16 |
+
ffmpy==0.3.0
|
| 17 |
+
filelock==3.12.2
|
| 18 |
+
fonttools==4.40.0
|
| 19 |
+
frozenlist==1.3.3
|
| 20 |
+
fsspec==2023.6.0
|
| 21 |
+
gradio==3.35.2
|
| 22 |
+
gradio_client==0.2.7
|
| 23 |
+
h11==0.14.0
|
| 24 |
+
httpcore==0.17.3
|
| 25 |
+
httpx==0.24.1
|
| 26 |
+
huggingface-hub==0.16.2
|
| 27 |
+
idna==3.4
|
| 28 |
+
Jinja2==3.1.2
|
| 29 |
+
jsonschema==4.17.3
|
| 30 |
+
kiwisolver==1.4.4
|
| 31 |
+
linkify-it-py==2.0.2
|
| 32 |
+
markdown-it-py==2.2.0
|
| 33 |
+
MarkupSafe==2.1.3
|
| 34 |
+
matplotlib==3.7.2
|
| 35 |
+
mdit-py-plugins==0.3.3
|
| 36 |
+
mdurl==0.1.2
|
| 37 |
+
multidict==6.0.4
|
| 38 |
+
numpy==1.25.0
|
| 39 |
+
orjson==3.9.1
|
| 40 |
+
packaging==23.1
|
| 41 |
+
pandas==2.0.3
|
| 42 |
+
Pillow==10.0.0
|
| 43 |
+
pydantic==1.10.11
|
| 44 |
+
pydub==0.25.1
|
| 45 |
+
Pygments==2.15.1
|
| 46 |
+
PyMuPDF==1.22.5
|
| 47 |
+
pyparsing==3.0.9
|
| 48 |
+
pyrsistent==0.19.3
|
| 49 |
+
pytesseract==0.3.10
|
| 50 |
+
python-dateutil==2.8.2
|
| 51 |
+
python-multipart==0.0.6
|
| 52 |
+
pytz==2023.3
|
| 53 |
+
PyYAML==6.0
|
| 54 |
+
requests==2.31.0
|
| 55 |
+
semantic-version==2.10.0
|
| 56 |
+
six==1.16.0
|
| 57 |
+
sniffio==1.3.0
|
| 58 |
+
starlette==0.27.0
|
| 59 |
+
toolz==0.12.0
|
| 60 |
+
tqdm==4.65.0
|
| 61 |
+
typing_extensions==4.7.1
|
| 62 |
+
tzdata==2023.3
|
| 63 |
+
uc-micro-py==1.0.2
|
| 64 |
+
urllib3==2.0.3
|
| 65 |
+
uvicorn==0.22.0
|
| 66 |
+
websockets==11.0.3
|
| 67 |
+
yarl==1.9.2
|