openfree's picture
Update app.py
1ec543a verified
import gradio as gr
import pandas as pd
from io import BytesIO
def convert_file(input_file, conversion_type):
# Check if a file was uploaded
if input_file is None:
return None, "Please upload a file."
# Read the file content
try:
# Try reading from file-like object
file_bytes = input_file.read()
file_name = input_file.name
except AttributeError:
# If there's an AttributeError, treat input_file as a file path
file_name = input_file
with open(file_name, "rb") as f:
file_bytes = f.read()
file_extension = file_name.lower().split('.')[-1]
df = None
output_file = None
converted_format = None
try:
# Conversion: CSV to Parquet
if conversion_type == "CSV to Parquet":
if file_extension != "csv":
return None, "For CSV to Parquet conversion, please upload a CSV file."
# Set UTF-8 as default encoding and try others if needed
encodings_to_try = ['utf-8', 'latin1', 'iso-8859-1', 'cp1252']
encoding = 'utf-8' # Set UTF-8 as default encoding
# Try UTF-8 first, then other encodings if it fails
try:
df = pd.read_csv(BytesIO(file_bytes), encoding=encoding)
except UnicodeDecodeError:
# Try other encodings if UTF-8 fails
for enc in encodings_to_try[1:]: # Try remaining encodings except utf-8
try:
df = pd.read_csv(BytesIO(file_bytes), encoding=enc)
encoding = enc
break
except UnicodeDecodeError:
continue
except Exception as e:
return None, f"Error reading CSV: {str(e)}"
except Exception as e:
return None, f"Error reading CSV: {str(e)}"
if df is None:
return None, "Failed to read CSV with any of the common encodings. Your file might use a rare encoding."
output_file = "output.parquet"
df.to_parquet(output_file, index=False)
converted_format = "Parquet"
# Conversion: Parquet to CSV
elif conversion_type == "Parquet to CSV":
if file_extension != "parquet":
return None, "For Parquet to CSV conversion, please upload a Parquet file."
df = pd.read_parquet(BytesIO(file_bytes))
output_file = "output.csv"
df.to_csv(output_file, index=False, encoding='utf-8')
converted_format = "CSV"
else:
return None, "Invalid conversion type selected."
# Generate a preview of the top 10 rows
preview = df.head(10).to_string(index=False)
info_message = (
f"Input file: {file_name}\n"
f"Converted file format: {converted_format}\n"
)
if conversion_type == "CSV to Parquet":
info_message += f"Used encoding: {encoding}\n"
info_message += f"\nPreview (Top 10 Rows):\n{preview}"
return output_file, info_message
except Exception as e:
return None, f"Error during conversion: {str(e)}"
# λͺ¨λ˜ν•˜κ³  μ„Έλ ¨λœ μŠ€νƒ€μΌμ„ μœ„ν•œ μ‚¬μš©μž μ •μ˜ CSS
custom_css = """
body {
background-color: #f4f4f4;
font-family: 'Helvetica Neue', Arial, sans-serif;
}
.gradio-container {
max-width: 900px;
margin: 40px auto;
padding: 20px;
background-color: #ffffff;
border-radius: 12px;
box-shadow: 0 8px 16px rgba(0,0,0,0.1);
}
h1, h2 {
color: #333333;
}
.gradio-input, .gradio-output {
margin-bottom: 20px;
}
.gradio-button {
background-color: #4CAF50 !important;
color: white !important;
border: none !important;
padding: 10px 20px !important;
font-size: 16px !important;
border-radius: 6px !important;
cursor: pointer;
}
.gradio-button:hover {
background-color: #45a049 !important;
}
"""
with gr.Blocks(css=custom_css, title="CSV <-> Parquet Converter") as demo:
gr.Markdown("# CSV <-> Parquet Converter")
gr.Markdown("Upload a CSV or Parquet file and select the conversion type. The app converts the file to the opposite format and displays a preview of the top 10 rows.")
with gr.Row():
with gr.Column(scale=1):
input_file = gr.File(label="Upload CSV or Parquet File")
with gr.Column(scale=1):
conversion_type = gr.Radio(
choices=["CSV to Parquet", "Parquet to CSV"],
label="Conversion Type",
value="CSV to Parquet" # Set default value
)
convert_button = gr.Button("Convert", elem_classes=["gradio-button"])
with gr.Row():
output_file = gr.File(label="Converted File")
preview = gr.Textbox(label="Preview (Top 10 Rows)", lines=15)
convert_button.click(fn=convert_file, inputs=[input_file, conversion_type], outputs=[output_file, preview])
gr.Markdown("""
### Notes:
- This converter uses UTF-8 as the default encoding
- If UTF-8 fails, it tries Latin-1, ISO-8859-1, and CP1252 encodings
- Parquet files preserve data types better than CSV
- The preview shows only the first 10 rows of data
""")
if __name__ == "__main__":
demo.launch()