Update LogtoClient.php

[digood-dev]

修复用户资料中姓名为中文名（例如“测试”两个字）时，响应的base64文本格式标准不一的问题
Fix the problem where the response base64 text format is inconsistent when the user's name in the profile is in Chinese.

[darcyYe]

Thanks for the fix. The base64url handling change makes sense, but two issues still need to be addressed before this is ready:

1. The fix only applies to getIdTokenClaims(). getAccessTokenClaims() still decodes the JWT payload with raw base64_decode(), so valid access tokens that require base64url normalization or padding will still fail. getOrganizationTokenClaims() inherits the same path.

2. The test suite still only covers ASCII-only payloads that already worked before this patch. Please add a regression test whose payload actually requires base64url normalization or padding, and cover both the ID-token and access-token claim helpers so this bug class cannot regress silently.

[darcyYe]

This only fixes base64url decoding for ID tokens. getAccessTokenClaims() still uses raw base64_decode() on the JWT payload, so valid access tokens that need URL-safe normalization or padding will still fail. Please extract this decode logic into a shared helper and reuse it for both token-claim paths.