fix: correct timeStrCache calculation to exclude unmatched trailing chars

[Snssn]

Use strptimeResult position (actual matched end) instead of curTimeStr.length()
when building the second-level cache. This fixes cache pollution when the time
string contains trailing characters not matched by the format string (e.g. timezone suffix +08:00)

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

senyong.wang seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[Takuka0311]

百炼自动化审查：建议保持开启。

本 PR 修复了 ProcessorParseTimestampNative 中时间字符串二级缓存键计算的缺陷，改用 strptime 实际匹配结束位置计算缓存长度以避免尾部未匹配字符导致的缓存污染，并补充了单元测试。因贡献者尚未签署 CLA 且等待维护者审查与 CI 验证，建议保持开启。.

最佳落地路径：

保持开启。等待贡献者完成 CLA 签署，维护者审查代码变更（重点验证 cacheLen 指针差值计算在各类边界条件下的安全性）及 CI 全量验证通过后合并至 main。.

已核对内容：

• pr_fix_logic: PR 修改缓存长度计算逻辑，使用 strptimeResult 指针差值获取实际匹配长度，解决尾部未匹配字符导致的缓存污染。. (core/plugin/processor/ProcessorParseTimestampNative.cpp:203, 5b5623946fd7)
• unit_test_added: 新增单元测试验证带尾部未匹配字符时间串的缓存行为，确保修复有效。. (core/unittest/processor/ProcessorParseTimestampNativeUnittest.cpp:894, 5b5623946fd7)
• cla_pending: CLA assistant 提示贡献者尚未签署 CLA，需等待签署完成后方可合并。.

百炼审查备注：模型 qwen3.6-max-preview；对照提交 8b91144f8a81。