fix(browser): handle config.base (#4686)

[mbland]

Description

Closes #4686.

Ensures that every import() is properly prefixed with either config.base or '/'.

Also extracted a runVitest() test helper to eliminate duplication in the new test. (Except that the new test file copies the first of the existing test cases.)

cc: @alexdulin

Please don't delete this checklist! Before submitting the PR, please make sure you do the following:

• It's really useful if your PR references an issue where it is discussed ahead of time. If the feature is substantial or introduces breaking changes without a discussion, PR might be closed.
• Ideally, include a test that fails without this PR but passes with it.
• Please, don't make changes to pnpm-lock.yaml unless you introduce a new test example.

Tests

• Run the tests with pnpm test:ci.

pnpm test passes locally on macOS; will try Linux and Windows 11 shortly. However, pnpm test:ci mostly passes, but eventually fails with:

test/vm-threads test:  ✓ test/module.test.js  (1 test) 3ms
test/vm-threads test: Failed
/Users/msb/src/mbland/vitest/test/vm-threads:
 ERR_PNPM_RECURSIVE_RUN_FIRST_FAIL  @vitest/test-vm-threads@ test: `vitest "--allowOnly"`
Exit status 129
test/workspaces test$ pnpm test:workspace && pnpm test:coverage "--allowOnly"
 ELIFECYCLE  Command failed with exit code 1.

I wonder if this might be due to the fact that I'm running on Apple Silicon, however. Any insight would be appreciated.

Documentation

• If you introduce new functionality, document it. You can run documentation with pnpm run docs command.

Changesets

• Changes in changelog are generated from PR name. Please, make sure that it explains your changes in an understandable manner. Please, prefix changeset messages with feat:, fix:, perf:, docs:, or chore:.

[netlify]

✅ Deploy Preview for fastidious-cascaron-4ded94 canceled.

Name	Link
🔨 Latest commit	f5f3251
🔍 Latest deploy log	https://app.netlify.com/sites/fastidious-cascaron-4ded94/deploys/6596add52de66b0008edac36

[mbland]

Hmm, the test ({macOS,windows}-latest) failures don't make a lot of sense to me yet. I don't see how this change could've affected them, and both pass on their respective OSes on my machine:

• https://github.com/vitest-dev/vitest/actions/runs/7123264063/job/19395610906?pr=4692#step:8:495

  test/filters test:  × test/testname-pattern.test.ts > match by partial pattern
  test/filters test:    → Test timed out in 5000ms.
  

• https://github.com/vitest-dev/vitest/actions/runs/7123264063/job/19395611206?pr=4692#step:10:828

  test/vite-node test:  ❯ test/cli.test.ts  (8 tests | 1 failed) 33619ms
  test/vite-node test:    ❯ test/cli.test.ts > correctly runs --watch index.cjs
  test/vite-node test:      → Timeout when waiting for output "test 2".
  

The test-browser{,-windows} (chrome, chromium) failures don't make sense yet, either, especially since the browser tests ran successfully on their respective OSes on my machine after seeing these failures:

• https://github.com/vitest-dev/vitest/actions/runs/7123264063/job/19395609661?pr=4692#step:10:22
• https://github.com/vitest-dev/vitest/actions/runs/7123264063/job/19395609801?pr=4692#step:9:21

I need to call it a night but can look more tomorrow.

[mbland]

After rebasing this morning to match main (tag v1.0.2), test (windows-latest, 20) and test-browser-windows (chrome, chromium) seem to be working again. test (macOS-latest) seems to be exhibiting the same snapshot failure I've seen in other builds, including the v1.0.2 commit on main:

• this PR: https://github.com/vitest-dev/vitest/actions/runs/7128065730/job/19409295347?pr=4692#step:8:855
• current main (v1.0.2): https://github.com/vitest-dev/vitest/actions/runs/7126712183/job/19405173121#step:10:749

test-browser (chrome, chromium) is more interesting. The test case I added passes, but now the existing tests fail:

• https://github.com/vitest-dev/vitest/actions/runs/7128065730/job/19409296285?pr=4692#step:10:21

What's more, several fail while spawning a WebDriver session due to ETXTBSY, which has an interesting history: lwn.net: The shrinking role of ETXTBSY.

Still looking to see how to account for this. I don't think it has to do with this change, but I'll try to make sure.

[mbland]

Ah, I think my new test suite is to blame; it's racing against the old one. Working on a fix.

[mbland]

The --test-concurrency=1 flag for test/browser seemed to do the trick! (Even test (macOS-latest) passed, though I don't think that had anything to do with this change.)

[sheremet-va]

Can you please resolve the conflict? I cannot force-push it.

[mbland]

@sheremet-va Done.