Windows环境下用Python将UTF-16格式纯ASCII文件转单字节ASCII

Got it, let's solve this problem quickly and cleanly. Since your input file uses UTF-16 encoding but only contains standard ASCII characters, converting it to a single-byte ASCII file is straightforward with Python—no need to modify the original file at all.

Step-by-Step Solution #

Here's a Python script tailored for Windows that handles the conversion safely:

# Set your input and output file paths (use raw strings to avoid Windows path escape issues)
input_path = r"C:\example\input_file.txt"
output_path = r"C:\example\output_file.txt"

try:
    # Read the UTF-16 encoded file
    with open(input_path, 'r', encoding='utf-16') as input_file:
        file_content = input_file.read()
    
    # Write the content as standard ASCII
    with open(output_path, 'w', encoding='ascii') as output_file:
        output_file.write(file_content)
    
    print(f"Conversion complete! Output saved to: {output_path}")
except FileNotFoundError:
    print(f"Error: Could not find input file at {input_path}")
except Exception as e:
    print(f"Unexpected error during conversion: {str(e)}")

Key Details: #

• Reading UTF-16: The encoding='utf-16' parameter tells Python to correctly parse the double-byte UTF-16 format, including handling the Byte Order Mark (BOM) that's common in Windows UTF-16 files.
• Writing ASCII: Since all characters are standard ASCII, using encoding='ascii' will write each character as a single byte, exactly what you need for a non-double-byte ASCII file.
• Windows Paths: Using raw strings (r"") for file paths prevents issues with backslash escape characters in Windows file paths.

Edge Case Note #

If for some reason there are hidden non-ASCII characters in the file (even though you mentioned all are ASCII), you can add an errors parameter to handle them. For example, encoding='ascii', errors='ignore' would skip non-ASCII characters, or errors='replace' would replace them with ?. But since you confirmed all characters are ASCII, this shouldn't be necessary.

内容的提问来源于stack exchange，提问作者Jeff Winchell