Skip to content

Codex/feature/case converter detected words - #418

Merged
anurag629 merged 2 commits into
codercops:developfrom
srikar-naidu:codex/feature/case-converter-detected-words
Oct 30, 2026
Merged

anurag629 merged 2 commits into
codercops:developfrom
srikar-naidu:codex/feature/case-converter-detected-words

Conversation

@srikar-naidu

Copy link
Copy Markdown
Contributor

What and why

Adds a “Detected words” section to the Text Case Converter.

It displays the word segments used by the existing conversion logic, preserving original casing and line associations. Supports acronyms, digits, separators, Unicode input, multiline text, and an empty state.

Closes #409

How I tested this

  • Ran npm run lint — passed with existing warnings only.
  • Ran npx tsc --noEmit — passed.
  • Added and reviewed unit tests for acronyms, digits, separators, Unicode, and multiline input.
  • npm run test and npm run build were attempted but blocked by environment-level Windows EPERM errors during Vitest cache renames and .next cleanup.

AI assistance

AI was used to implement the feature, add tests, review the diff, and run validation commands.

Type of change

  • Bug fix
  • New tool or feature
  • Tests
  • Refactor or cleanup
  • Docs

Checklist

  • I was assigned the linked issue
  • npm run lint && npm run test && npm run build passes locally
  • Logic changes live in lib/ and have a test in lib/__tests__/
  • For a bug fix: the new test fails without my change
  • No analytics, trackers, or calls that send user data off-device
  • Works in both light and dark themes (if UI changed)
  • Inputs have labels and it works with the keyboard alone (if UI changed)
  • New tool: registry entry, page.tsx, client component, opengraph-image.tsx, and the bug report dropdown are all done

Notes for reviewers

The existing splitWords conversion behavior is unchanged. Detection uses the same boundary rules through shared logic and exposes original-casing tokens for display.

The full test/build commands were blocked by sandbox Windows EPERM filesystem errors unrelated to this change.

@srikar-naidu
srikar-naidu force-pushed the codex/feature/case-converter-detected-words branch from 5b82613 to 1ef8b1b Compare October 29, 2026 02:05

@anurag629 anurag629 left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks @srikar-naidu. Lint (develop's 11 warnings), all 2444 tests, the build and the pages pass on top of develop, splitWords gives the same output as before, and detectWords matches the issue: XMLHttpRequest gives XML, Http, Request, HTTP_status_code gives HTTP, status, code, and blank lines are skipped with the original line numbers kept. One thing: the whole card has aria-live="polite", so a screen reader re-reads every chip on each keystroke while you type. Nothing else in the tool is a live region, so please drop it there.

@anurag629 anurag629 added the hacktoberfest-accepted PR counts for Hacktoberfest label Oct 30, 2026
@anurag629
anurag629 merged commit 8965911 into codercops:develop Oct 30, 2026
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

hacktoberfest-accepted PR counts for Hacktoberfest

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Text Case Converter: show how identifiers are split into words

2 participants