fix: improve misleading fetch() error for non-string URL inputs

[jw409]

Summary

Fixes #21361

When fetch() receives a non-string, non-URL value like undefined, 123, or an object whose toString() returns a relative path, Bun currently reports:

fetch() URL must not be a blank string.

This is misleading because the user did provide a value — it just wasn't a valid URL.

Changes

All changes are in src/bun.js/webcore/fetch.zig:

1. fetchImpl: Track whether the empty URL string came from actual user input (e.g. fetch("")) vs a failed type conversion (e.g. fetch(undefined)). When the input was a non-string/non-URL value that couldn't be converted, the error now reads:

   fetch() URL could not be parsed; received a non-string value that is not a URL object.

   The original "must not be a blank string" message is preserved for fetch("") where it is accurate.

2. Bun__fetchPreconnect_: When a URL parses successfully but has no hostname (e.g. a path like /some_path that made it past the empty check), the error now reads:

   fetch() URL must have a hostname.

   Previously this also used the "blank string" message, which was incorrect — the string wasn't blank, it just lacked a hostname.

Before / After

Input	Before	After
fetch(undefined)	URL must not be a blank string	URL could not be parsed; received a non-string value that is not a URL object
fetch(123)	URL must not be a blank string	URL could not be parsed; received a non-string value that is not a URL object
fetch("")	URL must not be a blank string	URL must not be a blank string (unchanged)
preconnect("/path")	URL must not be a blank string	URL must have a hostname

Test plan

• Verify fetch(undefined) no longer reports "blank string"
• Verify fetch(123) no longer reports "blank string"
• Verify fetch("") still reports "blank string" (correct case)
• Verify fetch({toString() { return '/path' }}) no longer reports "blank string"
• Verify fetch.preconnect('/path') reports "must have a hostname"
• Verify normal fetch('https://example.com') is unaffected

---

🇺🇸 Reid Wiseman — Commander
🇺🇸 Victor Glover — Pilot
🇺🇸 Christina Koch — Mission Specialist
🇨🇦 Jeremy Hansen — Mission Specialist

Artemis II. Open source for all — on this planet and beyond it.

[claude]

Claude Code Review

This pull request is from a fork — automated review is disabled. A repository maintainer can comment @claude review to run a one-time review.

[coderabbitai]

Walkthrough

Modified error handling in fetch.zig to provide more accurate error messages. Added fetch_error_no_hostname constant and updated Bun__fetchPreconnect_ to use it. Introduced url_was_from_input tracking in fetchImpl to distinguish between user-provided and derived URLs, enabling contextual error messages.

Changes

Cohort / File(s)	Summary
Fetch error handling improvements src/bun.js/webcore/fetch.zig	Added exported fetch_error_no_hostname constant and introduced url_was_from_input tracking variable in fetchImpl to distinguish user-provided URLs from derived ones. Updated Bun__fetchPreconnect_ to throw .INVALID_ARG_TYPE with the new error message when hostname is empty. Modified empty url_str error handling to provide context-specific messages based on URL origin.

🚥 Pre-merge checks | ✅ 4

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately summarizes the main change: improving misleading fetch() error messages for non-string URL inputs, which is the core objective.
Description check	✅ Passed	The description comprehensively covers both required sections: what the PR does and how to verify it, with detailed before/after examples and a complete test plan.
Linked Issues check	✅ Passed	The PR fully addresses issue #21361 by implementing more precise error messages for non-string URL inputs and missing hostname cases, matching the expected behavior.
Out of Scope Changes check	✅ Passed	All changes are scoped to error handling improvements in fetch.zig as specified in the linked issue, with no unrelated modifications.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@src/bun.js/webcore/fetch.zig`:
- Around line 342-349: Extract the inline string used when url_was_from_input is
false into a top-of-file constant (alongside existing error messages like
fetch_error_blank_url) and replace the inline literal in the conditional that
builds the TypeError (the branch that calls ctx.toTypeError(.INVALID_URL, "...",
.{})) so that the code uses the new constant instead of the inline string; keep
the same identifier naming style as other error constants and update references
in the url_str.isEmpty handling block and any other occurrences if present.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro

Run ID: 7fa69933-d199-4ce2-9244-ae5053647be9

📥 Commits

Reviewing files that changed from the base of the PR and between 1e178d1 and c40019a.

📒 Files selected for processing (1)

• src/bun.js/webcore/fetch.zig

[coderabbitai]

🧹 Nitpick | 🔵 Trivial

Consider extracting the inline error message to a constant for consistency.

The conditional error logic correctly addresses the PR objectives. However, the new error message on line 347 is defined inline rather than as a constant like the other error messages at the top of the file (lines 1-4).

🔧 Suggested refactor for consistency

Add a constant at the top of the file alongside other error messages:

 pub const fetch_error_no_args = "fetch() expects a string but received no arguments.";
 pub const fetch_error_blank_url = "fetch() URL must not be a blank string.";
 pub const fetch_error_no_hostname = "fetch() URL must have a hostname.";
+pub const fetch_error_non_string_url = "fetch() URL could not be parsed; received a non-string value that is not a URL object.";
 pub const fetch_error_unexpected_body = "fetch() request with GET/HEAD/OPTIONS method cannot have body.";

Then use it in the conditional:

     const err = if (url_was_from_input)
         ctx.toTypeError(.INVALID_URL, fetch_error_blank_url, .{})
     else
-        ctx.toTypeError(.INVALID_URL, "fetch() URL could not be parsed; received a non-string value that is not a URL object.", .{});
+        ctx.toTypeError(.INVALID_URL, fetch_error_non_string_url, .{});

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@src/bun.js/webcore/fetch.zig` around lines 342 - 349, Extract the inline
string used when url_was_from_input is false into a top-of-file constant
(alongside existing error messages like fetch_error_blank_url) and replace the
inline literal in the conditional that builds the TypeError (the branch that
calls ctx.toTypeError(.INVALID_URL, "...", .{})) so that the code uses the new
constant instead of the inline string; keep the same identifier naming style as
other error constants and update references in the url_str.isEmpty handling
block and any other occurrences if present.