[BUG]: Governance "Issue Resolution Rate" counts pull requests as issues

未关闭 适合新手
  • 哪里不对？
#398 1 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

维护者通常 4 天内回复

@amangupta-12 已经在做这个了。

开始于 2026年10月25日。

  • #413 来自 @amangupta-12 —— 未关闭

评估

难度
1/5
预计耗时
1 小时以内
新手友好度
86/100
Issue 类型
缺陷
描述清晰度
描述清楚
活跃度
活跃
技术栈
github, javascript, react
领域
analytics, frontend

调研方向

在 src/pages/GovernancePage.jsx 中，检查第338–343行附近的 Issue Resolution Rate 计算，并将其筛选逻辑与 issue 中提到的治理检查进行比较。从 repo issue 列表中排除 pull request，然后确认卡片的分母和比率只使用 issue；报告中的 OrgExplorer 示例应统计44个 issue，而不是获取到的100个项目。

由索引模型根据 Issue 内容生成。

描述

bug
Bug Description

The Governance "Issue Resolution Rate" card computes each repo's rate from GitHub's
/issues data without excluding pull requests. GitHub's /issues?state=all returns PRs as
well as issues (every PR is also an issue object, identified by a pull_request key), so
merged/closed PRs are counted as resolved issues and the rate is wrong.

In src/pages/GovernancePage.jsx (~lines 338-343):

const repoIssues = allIssues.filter(i => i.repoName === r.name)
const closed = repoIssues.filter(i => i.state === 'closed').length
const total = repoIssues.length
const rate = total ? Math.round(closed / total * 100) : null

allIssues (built ~lines 107-114) includes PRs and is never filtered with !i.pull_request.
Every other governance check DOES filter PRs out — Dead Issues (line 126), Zombie PRs
(line 144), and the stale-issues stats (AppContext.jsx line 422) — so this widget is the
outlier.

Verified on live data for AOSSIE-Org/OrgExplorer (first 100 of state=all): 56 of the 100
"issues" are actually PRs, so the denominator is inflated 44 -> 100 and the rate reads 42%
instead of the correct 39%.

Steps to Reproduce
  1. Explore any active org (e.g. AOSSIE-Org) and open the Governance page.
  2. Click "Run Audit".
  3. Look at the "Issue Resolution Rate" bars — closed/total includes pull requests.
  4. Compare any repo's number with GitHub's real issue count and it won't match.

Deterministic check (same logic as the page, on live data):

node f1_resolution.mjs
# fetched items (issues + PRs): 100  (PRs among them: 56)
# CURRENT (buggy, counts PRs) : 42/100 closed -> 42%
# CORRECT (issues only)       : 17/44  closed -> 39%
Logs and Screenshots

Reproducible script output (replicates GovernancePage.jsx:338-343 on live GitHub data):

=== F1 Issue Resolution Rate — AOSSIE-Org/OrgExplorer (first 100 of state=all) ===
fetched items (issues + PRs): 100 (PRs among them: 56)

CURRENT (buggy, counts PRs) : 42/100 closed -> 42%
CORRECT (issues only) : 17/44 closed -> 39%

=> PRs inflate the denominator by 56 and skew the rate by 3 points.

📎 Image file:

Image
Environment Details
  • OS: macOS (also reproduces on the live site, any OS/browser)
  • Browser: Chrome
  • Site: https://orgexplorer.aossie.org/ and local dev (npm run dev), main @ d91d2b8
  • Node.js: v20 (for the repro script)
  • Suggested fix: add && !i.pull_request to the repoIssues filter.
  • Related but separate: #302's open PR #304 rewrites this exact line for org-scoping
    (issuesData?.[${r.orgLogin}/${r.name}] || []) but still counts PRs — so this defect
    survives #304 and should be fixed on top of / within that change.
Impact

Medium - Feature works but has issues

Code of Conduct
  • I have joined the Discord server and will post updates there
  • I have searched existing issues to avoid duplicates
主要语言
JavaScript
星标
48
派生
169
平均合并
4 天 13 小时
30 天内合并 PR
25

环境准备

从这里开始

  1. 先读完整个 Issue，再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库，在一个分支上完成修改。
  4. 提交 Pull Request，并在描述里引用这个 Issue 编号。

AOSSIE-Org/OrgExplorer 的其他 Issue

查看 AOSSIE-Org/OrgExplorer 的全部 Issue

相似的 Issue

更多 JavaScript Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。