Skip to content

chore:Feat/off input and tag filter string - #7812

Open
SteelHe wants to merge 2 commits into
labring:mainfrom
SteelHe:feat/off-input-and-tag-filter-string
Open

SteelHe wants to merge 2 commits into
labring:mainfrom
SteelHe:feat/off-input-and-tag-filter-string

Conversation

@SteelHe

@SteelHe SteelHe commented Sep 22, 2026

Copy link
Copy Markdown
Contributor

summary

完全不影响fastgpt 的前端和后台

  1. FlowNodeInputTypeEnum 增加off类型
  2. 标签过滤DatasetCollectionTagTypeEnum增加string,但是页面不渲染
  3. 非fastgpt格式的标签过滤,增加对变量引用的计算,后续升级的兜底

validate

  1. 单测
  2. 页面端到端测试

SteelHe and others added 2 commits September 22, 2026 07:54
off 只给第三方开发者接口使用:页面不渲染该输入,但元数据、默认值和 schema
校验都保留,接口仍可正常传参。

- FlowNodeInputMap / RenderList 补 off;filterInputs 里与 hidden 一起跳过渲染
  (只把 RenderList[off] 设成 undefined 仍会渲染出行标签)
- jsonschema 的 workflowToolPreservedInputRenderTypes 保留 off,否则
  workflow tool 往返会丢 renderTypeList / selectedType
- 通用必填校验跳过 hidden / off,避免 UI 不可见的 required 输入卡住发布
- InputEditModal 不为 off 覆盖 valueType:它不在类型选择器里,会拿到兜底值

Co-Authored-By: Claude <noreply@anthropic.com>
- WorkflowTagFilterTagTypeSchema 增加 string,接口下发的结构化 string 条件
  能被 isDatasetTagFilterValue 识别,并交给检索层执行(checkValue 的 string
  分支已支持 $eq/$ne/$contains/$notContains/$startsWith/$endsWith/$regex/
  $empty/$notEmpty)
- 标签下拉仍不出现 string(isInterfaceOnlyTagFilterTagType),条件行照常渲染,
  补上文本输入和引用变量类型过滤;i18n 补 开头是/结尾是/模糊匹配
- pruneTagFilterConditions 放行 string:界面没有创建入口,也就无权按下拉候选
  判定它失效,否则已选库一变就会被悄悄删掉
- formatCollectionFilterMatchParam 解引用 JSON 字符串里内嵌的
  ['$ref', nodeId, outputKey](引用嵌在字符串内部,getReferenceVariableValue
  只认顶层引用);解不出时原样保留,避免产出检索层会拒的残缺条件对象

Co-Authored-By: Claude <noreply@anthropic.com>

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant