Removing URLs and Emails Before Sharing Text
A privacy-scrub workflow with boundary tests and a checklist for what automation cannot find.
Tested 2026-07-18 · Method: 20 tokens including punctuation, subdomains, email aliases, and false positives · By Dionis Ceban
Bottom line: The safest workflow makes normalization choices explicit, checks a small sample first, and keeps a human review step for meaning-sensitive changes.
Test matrix
| Input or approach | Operation | Observed result |
|---|---|---|
| https://example.com/report?user=7 | URL removal | Removed |
| name+sales@example.co.uk | Email removal | Removed |
| example.com without scheme | URL removal | May remain; review required |
| User ID 73921 | URL + email removal | Remains; not an email or URL |
Recommended workflow
- Run URL and email removal.
- Search the result for @, http, www, and common domain endings.
- Review names, phone numbers, account IDs, and addresses manually.
- Use a redaction process—not this convenience tool—for regulated or high-risk data.
How to reproduce this test
Copy the examples from the matrix into the relevant tool, apply only the named setting, and compare the output character by character. Browser and operating-system differences should not affect these JavaScript text operations.
What this test does not prove
This is a focused behavior test, not a guarantee for every document. Language rules, proprietary file formats, personally identifying data, and search-engine rendering all require context-specific review.