Skip to content

Releases: nowscott/CheckinQualityTool

v2.10.12

Choose a tag to compare

@nowscott nowscott released this 30 Sep 03:11
f97fdb5

主管和助理主管仍可参加普通抽检,但 是否生成报告=否 的课程不再容量外加抽。经理仍从所有抽检中排除。规则同时应用于上传的在职名单和项目内置快照,并在处理说明与页面反馈中显示跳过的课程数。

已通过 21 项抽检回归检查、类型检查、API 类型检查和生产构建。

v2.10.11

Choose a tag to compare

@nowscott nowscott released this 30 Sep 02:42
678f914

更新内容

  • 抽检名单默认只含教师姓名、学员姓名、课次日期、课次时间。
  • “入选原因”改为独立可选列,默认不导出。
  • “处理说明”继续独立可选;“未生成报告风险”页继续移除。

验证

  • 抽检检查 20 项通过,分别核对原因列开关和说明页开关。
  • 类型检查、API 类型检查及生产构建通过。

v2.10.10

Choose a tag to compare

@nowscott nowscott released this 30 Sep 02:38
3eb76fc

更新内容

  • 在“抽检名单”的 15 个追溯字段前恢复基础信息:教师姓名、学员姓名、课次日期、课次时间。
  • “未生成报告风险”页仍移除;“处理说明”仍可选输出。

验证

  • 抽检检查 19 项通过,包含表头顺序与示例数据核对。
  • 类型检查、API 类型检查和生产构建通过。

v2.10.9

Choose a tag to compare

@nowscott nowscott released this 30 Sep 02:31
da77f91

更新内容

  • 恢复“抽检名单”页的 15 个字段:抽检序号、抽检类型、入选原因、业务周、抽取尝试次数、源表行号、课次 ID、教师邮箱、报告链接 H5、报告生成状态、报告生成时间、产品分组、项目组、校区、科目。
  • 不额外加入手动模板列;“未生成报告风险”页仍移除,“处理说明”页仍可选输出。

验证

  • 抽检检查 19 项通过,包含 Excel 表头顺序和示例行核对。
  • 类型检查、API 类型检查与生产构建通过。

v2.10.8

Choose a tag to compare

@nowscott nowscott released this 30 Sep 02:28
4d9c13a

更新内容

  • “抽检名单”页只保留“入选原因”一列;抽检序号、类型、业务周、课程与教师追溯字段全部移除。
  • “未生成报告风险”页继续移除;“处理说明”页继续可选输出,默认关闭。

验证

  • 抽检相关检查 19 项通过,包含 Excel 表头和行数据断言。
  • 类型检查与生产构建通过。

v2.10.7

Choose a tag to compare

@nowscott nowscott released this 23 Sep 06:51
8df85d8

修复

  • 唯一教师姓名直接匹配评分,不再因教研组或师训组长字段变化而漏配。
  • 仅当教师姓名重名时,使用教研组和师训组长消歧。

验证

  • 抽检相关检查 19 项通过。
  • 2026-09-22 在职表模拟匹配:894 位候选教师中 871 位匹配评分,23 位无匹配,0 位身份歧义;6 位经理排除。
  • TypeScript 类型检查、API 类型检查、生产构建通过。

v2.10.6

Choose a tag to compare

@nowscott nowscott released this 23 Sep 06:51
7034a6c

更新内容

  • 接入腾讯文档“FY27Q1教学服务”页赋分排序;服务端仅返回教师身份字段和优先顺序,不返回原始分数。
  • 抽检名单改为指定的 15 列;移除“未生成报告风险”页;“处理说明”可选,默认关闭。
  • 保留经理排除、主管参与及现有两种优先级逻辑。

验证

  • 抽检选择与 Excel 页签/列结构检查通过。
  • TypeScript 类型检查、API 类型检查、生产构建通过。

v2.10.5

Choose a tag to compare

@nowscott nowscott released this 23 Sep 06:08
2d004b8

Changes

  • Keep all report-not-generated courses as capacity-exempt extra samples.
  • Coverage priority: cover teachers first; if the limit is short, preserve focus-list teachers and choose unscored/low-scored teachers before higher-scored non-focus teachers. Once everyone is covered, extra ordinary samples prioritize focus teachers, then lower scores.
  • Unreported priority: focus-list teachers first; with insufficient capacity, prioritize unscored/lower-scored non-focus teachers. After initial coverage, ordinary extras follow low-score order.
  • Add an optional teacher-email-to-score input through the worker request. The score source is not integrated yet, so current exports continue to use stable ordering when scores are absent.

Validation

  • TypeScript typecheck and production build passed.
  • All 13 inspection sampler checks passed, including score-aware shortage/extra ordering and protection for focus-list teachers.
  • git diff --check passed.

v2.10.4

Choose a tag to compare

@nowscott nowscott released this 23 Sep 05:36
99b381a

Changes

  • Teacher-coverage mode now samples one course per teacher first, then uses remaining ordinary sample slots for extra courses from teachers on the monthly unreported list.
  • If the focus list cannot be read, coverage mode still exports and fills remaining slots in stable order; unreported-priority mode continues to require the list.
  • Export reasons and processing notes now show unreported-teacher extra-course counts.

Validation

  • TypeScript typecheck and production build passed.
  • All 9 inspection sampler checks passed, including coverage followed by focus-teacher frequency allocation.
  • git diff --check passed.

v2.10.3

Choose a tag to compare

@nowscott nowscott released this 23 Sep 05:24
8bf7f06

Changes

  • Exclude teachers whose role description contains 经理 from inspection sampling.
  • Keep 主管 and 助理主管 teachers eligible for sampling.
  • Regenerate the built-in role exclusion snapshot and roster-generation logic to include only manager roles.

Validation

  • TypeScript typecheck and production build passed.
  • All 8 inspection checks passed, including manager-only exclusion.
  • The 2026-09-22 roster has 899 valid emails; its manager-only exclusion snapshot contains 6 emails.
  • Confirmed supervisor entries remain on the roster and are not excluded.