大文件Base64编解码异常求助：出现乱码字符

It sounds like your chunked approach is causing invalid Base64 data—specifically because per-chunk padding characters (=) are being inserted in the middle of your encoded string, which breaks the Base64 specification (padding should only appear at the very end of the entire encoded content). Here's why that happens and how to fix it:

Root Cause #

Base64 encodes 3 bytes of binary data into 4 ASCII characters. If a chunk isn’t a multiple of 3 bytes, base64_encode() will add padding (= or ==) to that chunk. When you concatenate these chunked encodings, those middle padding characters make the entire string invalid, leading to garbled data during decoding.

Corrected Encoding Code #

Instead of using arbitrary chunk sizes, use a chunk size that’s a multiple of 3 bytes (so no padding is added mid-stream). Only the final chunk (which may be smaller than 3 bytes) will get padding, which is valid:

function encode_file($inputFilename, $outputFilename) {
    $handle = fopen($inputFilename, 'rb');
    if (!$handle) {
        throw new Exception("Failed to open input file");
    }
    
    $outHandle = fopen($outputFilename, 'wb');
    if (!$outHandle) {
        fclose($handle);
        throw new Exception("Failed to open output file");
    }
    
    // Use a chunk size that's a multiple of 3 (e.g., 3072 bytes = 3*1024)
    $chunkSize = 3 * 1024;
    
    while (!feof($handle)) {
        $chunk = fread($handle, $chunkSize);
        $encodedChunk = base64_encode($chunk);
        fwrite($outHandle, $encodedChunk);
    }
    
    fclose($handle);
    fclose($outHandle);
}

Corrected Decoding Code #

For decoding, use a chunk size that’s a multiple of 4 bytes (since each 4 Base64 characters decode to 3 bytes). Also, ensure you handle the final chunk (which may include padding) correctly:

function decode_file($inputFilename, $outputFilename) {
    $handle = fopen($inputFilename, 'rb');
    if (!$handle) {
        throw new Exception("Failed to open encoded input file");
    }
    
    $outHandle = fopen($outputFilename, 'wb');
    if (!$outHandle) {
        fclose($handle);
        throw new Exception("Failed to open output file");
    }
    
    // Chunk size is multiple of 4 (e.g., 4096 bytes = 4*1024)
    $chunkSize = 4 * 1024;
    
    while (!feof($handle)) {
        $chunk = fread($handle, $chunkSize);
        // Remove any whitespace (if your encoded file has line breaks)
        $chunk = preg_replace('/\s+/', '', $chunk);
        $decodedChunk = base64_decode($chunk, true); // Strict error checking
        
        if ($decodedChunk === false) {
            fclose($handle);
            fclose($outHandle);
            throw new Exception("Invalid Base64 data encountered");
        }
        
        fwrite($outHandle, $decodedChunk);
    }
    
    fclose($handle);
    fclose($outHandle);
}

Key Tips #

• Chunk Size Alignment: Stick to chunk sizes divisible by 3 (encoding) or 4 (decoding) to avoid mid-stream padding issues.
• Strict Decoding: Using base64_decode($chunk, true) returns false if invalid Base64 is found, making debugging easier.
• Binary Mode: You’re already using 'rb' and 'wb' modes—this is critical to prevent line-ending conversion that could corrupt binary data.
• Whitespace Handling: If your encoded file has line breaks (common in some Base64 implementations), strip them before decoding with the regex shown.

内容的提问来源于stack exchange，提问作者a4w