i-code v0.1.4 #3
xucux
announced in
Announcements
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
[0.1.4] - 2026-08-05
新增
***),位于「模型 ID」下方一行,随导出写入 CSV / JSONrequest_headers_to_json去敏序列化(logging/headers.rs):将HeaderMap序列化为有序 JSON;头名称(不区分大小写、子串匹配)命中authorization / api-key / token / secret / credential / cookie / auth任一敏感片段时值替换为***;非 UTF-8 二进制值序列化为<binary>;无请求头时返回NoneUpstreamClient::execute签名改为&mut UpstreamContext,各 Client(openai_chat / anthropic / openai_responses / websocket)在发送前对真实出站请求头做去敏快照写入UpstreamContext.request_headers_json,供转发(provider-api)日志展示chat/responses/messages/models)捕获入站HeaderMap并透传到网关日志与转发日志LogRecord/LogEntry新增request_headers: Option<String>字段;CSV 导出新增requestHeaders列LogViewer以 JSON 缩进格式展示请求头,i18n(zh-CN / en / ja / zh-TW)补充requestHeaders文案i_code::sse写入单独按小时滚动的i-code-sse.*.log(前缀i-code-sse),不混入主日志文件,避免高频 chunk 刷屏TraceIdFormat::without_location(),文件内不打印 target 与 file:lineMainLogFilter)排除i_code::ssetarget,确保 chunk 只进专属文件变更
SizeAwareFileAppender):写入超限时不只判断滚动,改为按max_size拆分 buffer——先写满当前文件、滚动到新分段后继续写剩余部分,保证每个分段不超过max_size(修复单块大写入出现 ~40MB 分段的隐患);使用create_new(O_EXCL)独占创建分段并跳过已被占用的序号,从根上避免多进程 / 多次重启把同一分段追加撑大或序号复用修复
/v1路径:网关 API 文档弹窗与网关首页的「复制 API 地址」按钮此前只复制http://host:port,现统一补充/v1,复制完整的 OpenAI 兼容基础地址This discussion was created from the release i-code v0.1.4.
All reactions