fix(transform): use scheme as domain for file:// URLs

[TimeToBuildBob]

Summary

• When split_url_events processes file://, about:, or other URLs without a netloc, it now uses the URL scheme as $domain instead of an empty string
• This prevents all non-web URLs from silently clustering together as a blank entry in "Top Browser Domains"
• For example, file:///home/user/doc.pdf now gets $domain = "file" instead of $domain = ""

Test plan

• Updated existing file:// test case to expect $domain == "file" instead of ""
• Added test case for about:blank (expects $domain == "about")
• All 156 tests pass (2 skipped, pre-existing)

Fixes #67

---

Important

split_url_events() now uses the URL scheme as $domain for URLs without a netloc, preventing clustering as a blank entry.

• Behavior:
  • In split_url_events.py, split_url_events() now uses the URL scheme as $domain for URLs without a netloc (e.g., file://, about:).
  • Prevents non-web URLs from clustering as a blank entry in "Top Browser Domains".
• Tests:
  • Updated test_url_parse_event() in test_transforms.py to expect $domain == "file" for file:// URLs.
  • Added test case for about:blank to expect $domain == "about".
  • All tests pass, ensuring correct behavior for URLs without a netloc.

^{This description was created by for a4dd1e3. You can customize this summary. It will automatically update as commits are pushed.}

[ellipsis-dev]

Important

Looks good to me! 👍

Reviewed everything up to a4dd1e3 in 13 seconds. Click for details.

• Reviewed 54 lines of code in 2 files
• Skipped 0 files when reviewing.
• Skipped posting 0 draft comments. View those below.
• Modify your settings and rules to customize what types of comments Ellipsis leaves. And don't forget to react with 👍 or 👎 to teach Ellipsis.

Workflow ID: wflow_j7JlFqI4J6cLqwhv

^{You can customize by changing your verbosity settings, reacting with 👍 or 👎, replying to comments, or adding code review rules.}

[greptile-apps]

Greptile Summary

This PR fixes how non-web URLs (file://, about:, etc.) are handled in the URL parser by using their scheme as the $domain value instead of an empty string. Previously, all URLs without a netloc would get an empty domain, causing them to cluster together in "Top Browser Domains" visualizations. Now each scheme type gets its own identifiable domain (e.g., file:///path → $domain = "file", about:blank → $domain = "about").

The fix is backward compatible - regular HTTP/HTTPS URLs with netloc continue to work exactly as before, including the existing "www." prefix stripping logic. The change only affects URLs that previously had empty domains, making them more useful for analytics and reporting.

Confidence Score: 5/5

• This PR is safe to merge with minimal risk
• The change is well-tested (existing tests updated, new test added), backward compatible for web URLs, and solves a real issue. The logic is straightforward with proper fallback handling, and all 156 tests pass.
• No files require special attention

Important Files Changed

Filename	Overview
aw_transform/split_url_events.py	Adds conditional logic to use scheme as $domain for URLs without netloc (file://, about:, etc.), preventing empty domain clustering
tests/test_transforms.py	Updates file:// test expectation and adds about:blank test case to verify scheme-as-domain behavior

_{Last reviewed commit: a4dd1e3}

[greptile-apps]

_{2 files reviewed, no comments}

_{Edit Code Review Agent Settings | Greptile}