Run & review this pull request in StackBlitz Codeflow.

Reviewed PR #3686. The regex simplification from .replace(/\r\n/g, '\n').replace(/\r/g, '\n') to .replace(/\r\n?/g, '\n') is semantically equivalent and correct. The remaining changes (arrow→function declaration, style normalization in tests, 280+ matching snapshot updates) are all consistent. No issues found.

  ^{｜ View workflow run ｜ Triggered by Pullfrog ｜ 𝕏}

🦋 Changeset detected

Latest commit: c9c1d44

The changes in this PR will be included in the next version bump.

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
hey-api-docs	Ready	Preview, Comment	Apr 5, 2026 11:28pm

TL;DR — Simplifies SSE line-ending normalization from two chained .replace() calls to a single regex (/\r\n?/g), and converts createSseClient from an arrow function to a function declaration. Applied across all client plugins and regenerated snapshots.

Key changes

• Simplify SSE line-ending normalization to a single regex — Replaces the two-pass replace(/\r\n/g, '\n').replace(/\r/g, '\n') with the equivalent single-pass replace(/\r\n?/g, '\n') in the serverSentEvents.ts bundle.
• Convert createSseClient to a function declaration — Changes from export const createSseClient = <TData>(...) => { to export function createSseClient<TData>(...) {, aligning with standard function declaration style.
• Minor lint fixes in tests — Updates any[] to Array<any> and string[] to ReadonlyArray<string> in serverSentEvents.test.ts.
• Regenerated snapshots and examples — All 270+ serverSentEvents.gen.ts snapshot and example files updated to reflect the source changes.

_{Summary ｜ 283 files ｜ 1 commit ｜ base: main ← refactor/sse-line-endings}

This is a single-feature PR. The two-pass approach (\r\n → \n, then \r → \n) and the single-pass regex (\r\n? → \n) are functionally identical — \r\n? matches \r\n (CRLF) first when followed by \n, and bare \r (CR) otherwise, covering all three line-ending styles in one pass.

Before: buffer = buffer.replace(/\r\n/g, '\n').replace(/\r/g, '\n'); with a separate comment line
After: buffer = buffer.replace(/\r\n?/g, '\n'); with an inline comment

Why does /\r\n?/g work as a single-pass replacement?

The regex \r\n? greedily matches \r followed by an optional \n. When the input contains \r\n (CRLF), the engine matches both characters and replaces them with \n. When it encounters a bare \r (old Mac-style CR), it matches just the \r and replaces it with \n. Lone \n characters are unaffected since \r is required. This is a well-known single-regex idiom for line-ending normalization.

serverSentEvents.ts · serverSentEvents.test.ts

  ^{｜ View workflow run ｜ Triggered by Pullfrog ｜ 𝕏}

Open in StackBlitz

npm i https://pkg.pr.new/@hey-api/codegen-core@3686

npm i https://pkg.pr.new/@hey-api/json-schema-ref-parser@3686

npm i https://pkg.pr.new/@hey-api/nuxt@3686

npm i https://pkg.pr.new/@hey-api/openapi-ts@3686

npm i https://pkg.pr.new/@hey-api/shared@3686

npm i https://pkg.pr.new/@hey-api/spec-types@3686

npm i https://pkg.pr.new/@hey-api/types@3686

npm i https://pkg.pr.new/@hey-api/vite-plugin@3686

commit: c9c1d44

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 39.39%. Comparing base (8d3167c) to head (c9c1d44).
⚠️ Report is 2 commits behind head on main.

@@            Coverage Diff             @@
##             main    #3686      +/-   ##
==========================================
- Coverage   39.39%   39.39%   -0.01%     
==========================================
  Files         520      520              
  Lines       19279    19278       -1     
  Branches     5714     5708       -6     
==========================================
- Hits         7595     7594       -1     
  Misses       9445     9445              
  Partials     2239     2239              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.