Intro

Intro

Sometimes you encounter a base64 string but you don’t know the original text encoding. This guide shows a Python script that:

• Sanitizes the base64 input,
• Fixes missing padding,
• Decodes the bytes,
• Attempts to convert the bytes to text using several likely encodings.

import base64
# Put your base64 string here (can be multi-line)
base64_string = """<paste your base64 string here>"""
# Remove whitespace/newlines
base64_string = ''.join(base64_string.split())
# Fix missing padding (base64 strings length must be multiple of 4)
missing_padding = len(base64_string) % 4
if missing_padding:
    base64_string += '=' * (4 - missing_padding)
# Decode base64 to raw bytes
decoded_bytes = base64.b64decode(base64_string)
# Try a list of common encodings until one succeeds
encodings = ['utf-16le', 'utf-16', 'utf-8', 'latin-1']
for encoding in encodings:
    try:
        decoded_string = decoded_bytes.decode(encoding, errors='ignore')
        print(f"Successfully decoded with {encoding}:\n")
        print(decoded_string)
        break
    except Exception as e:
        print(f"Failed with {encoding}: {e}")

Explanation

import base64

• Imports Python’s standard library for base64 encoding/decoding.

base64_string = “””…”””

• Place your base64 string here. Using triple quotes lets you paste multi-line strings directly.

base64_string = ‘’.join(base64_string.split())

• Remove whitespace and newlines that might have been introduced by copy/paste or line-wrapping.

missing_padding = len(base64_string) % 4

• Base64 encodes data in 4-character blocks. If the length isn’t a multiple of 4, padding using ‘=’ is required.

if missing_padding: base64_string += ‘=’ * (4 — missing_padding)

• Append the correct number of ‘=’ characters to restore valid padding.

decoded_bytes = base64.b64decode(base64_string)

• Convert base64 to raw bytes. This can raise an exception if the string is invalid.

encodings = [‘utf-16le’, ‘utf-16’, ‘utf-8’, ‘latin-1’]

• A small prioritized list of encodings to try. Order is chosen because many Windows/UTF-16 variants appear in such cases; adjust as needed.

decoded_bytes.decode(encoding, errors=’ignore’)

• Attempt to decode bytes to text. errors=’ignore’ will drop invalid byte sequences rather than raising; you can use errors=’replace’ to get placeholders for invalid bytes.
• The loop stops when one encoding yields text; otherwise you’ll see failure messages for each tried encoding.

Why the encoding list and order?

• utf-16le / utf-16: Byte-order-marked or little-endian UTF-16 is common for Windows-generated text or some exports.
• utf-8: The most common modern text encoding.
• latin-1: Single-byte mapping that will always succeed (no decode errors) and preserves raw byte values 0–255 — useful as a last resort if you want to inspect raw characters.

Safety and caveats

• errors=’ignore’ hides decoding problems; if you want to spot issues, use errors=’strict’ (raises exceptions) or errors=’replace’
• base64 input might represent binary files (images, executables). Don’t assume textual content — inspect the bytes (e.g., check magic/header bytes) before treating them as text.
• Avoid decoding untrusted data in insecure environments. If the decoded content is saved and executed, it could be harmful.

Optimise my code

• Auto-detect encoding with chardet or charset-normalizer:

pip install charset-normalizer

• Use charset_normalizer from_bytes(decoded_bytes).best() to suggest encodings.
• Save bytes to a file if the content might be binary:

with open(‘output.bin’, ‘wb’) as f: f.write(decoded_bytes)

• Print a small bytes hex preview to inspect binary signatures:

print(decoded_bytes[:32].hex())

Example: using charset-normalizer (small snippet)

from charset_normalizer import from_bytes
result = from_bytes(decoded_bytes).best()
if result:
    print("Detected encoding:", result.encoding)
    print(result.first().decoded)

How to run

• Save the script to decode_base64.py and run: python decode_base64.py (Python 3.8+ recommended).