短網址服務核心需求:長整數ID ↔ 短字符串碼雙向無衝突轉換。本文實現基於Base54(排除易混淆字符)的編解碼算法,分析衝突風險、擴展性及生產環境優化。
排除0/O、1/I/l等易混淆字符,保留54字符:
$codes = 'abcdefghjkmnpqrstuvwxyz23456789ABCDEFGHJKMNPQRSTUVWXYZ'; // len=54
function generate_code(int $number): string {
if ($number < 0) throw new InvalidArgumentException('僅支持非負整數');
$out = '';
$codes = 'abcdefghjkmnpqrstuvwxyz23456789ABCDEFGHJKMNPQRSTUVWXYZ';
$base = strlen($codes); // 54
do {
$key = $number % $base;
$number = intdiv($number, $base);
$out = $codes[$key] . $out;
} while ($number > 0);
return $out;
}
// 測試
echo generate_code(123456789); // 輸出如 "aB3k9"
function get_num(string $code): int {
$codes = 'abcdefghjkmnpqrstuvwxyz23456789ABCDEFGHJKMNPQRSTUVWXYZ';
$base = strlen($codes);
$num = 0;
foreach (str_split($code) as $char) {
$pos = strpos($codes, $char);
if ($pos === false) throw new InvalidArgumentException("非法字符: $char");
$num = $num * $base + $pos;
}
return $num;
}
// 驗證
$num = 123456789;
$code = generate_code($num);
assert(get_num($code) === $num); // 通過
0輸出$codes[0]而非空字符串。$number = floor($number / 54) - 1邏輯有誤,導致編碼非連續、解碼失敗;標準BaseN無需減1。9.22e18),超大ID需用gmp或bcmath擴展。u_用戶、p_商品、o_訂單。code → id,減少資料庫查詢。正確的Base54實現配合分佈式ID生成,可構建高性能、無衝突的短網址服務。