feat: support watch protocol v1

[jbedard]

---

Changes are visible to end-users: yes/no

• Searched for relevant documentation and updated as needed: yes/no
• Breaking change (forces users to change their own code or config): yes/no
• Suggested release notes appear below: yes/no

Test plan

• Covered by existing test cases
• New test cases added
• Manual testing; please provide instructions so we can reproduce:

[aspect-workflows]

Test

⚠️ Buildkite build #9749 failed.

Failed tests (2)

//js/private/watch:watch_checked_0_test [k8-fastbuild] 🔗
//js/private/watch:watch_checked_1_test [k8-fastbuild] 🔗

💡 To reproduce the test failures, run

bazel test //js/private/watch:watch_checked_0_test //js/private/watch:watch_checked_1_test

---

Test

e2e/bzlmod

All tests were cache hits

5 tests (100.0%) were fully cached saving 618ms.

---

Test

e2e/git_dep_metadata

All tests were cache hits

1 test (100.0%) was fully cached saving 31ms.

---

Test

e2e/gyp_no_install_script

All tests were cache hits

2 tests (100.0%) were fully cached saving 173ms.

---

Test

e2e/js_image_oci

All tests were cache hits

1 test (100.0%) was fully cached saving 7s.

---

Test

e2e/npm_link_package

All tests were cache hits

3 tests (100.0%) were fully cached saving 407ms.

---

Test

e2e/npm_link_package-esm

All tests were cache hits

3 tests (100.0%) were fully cached saving 484ms.

---

Test

e2e/npm_link_package-rerooted

All tests were cache hits

3 tests (100.0%) were fully cached saving 536ms.

---

Test

e2e/npm_translate_lock

All tests were cache hits

3 tests (100.0%) were fully cached saving 919ms.

---

Test

e2e/npm_translate_lock_disable_hooks

All tests were cache hits

3 tests (100.0%) were fully cached saving 247ms.

---

Test

e2e/npm_translate_lock_empty

All tests were cache hits

1 test (100.0%) was fully cached saving 31ms.

---

Test

e2e/npm_translate_lock_exclude_package_contents

All tests were cache hits

1 test (100.0%) was fully cached saving 34ms.

---

Test

e2e/npm_translate_lock_multi

All tests were cache hits

2 tests (100.0%) were fully cached saving 164ms.

---

Test

e2e/npm_translate_lock_partial_clone

All tests were cache hits

1 test (100.0%) was fully cached saving 61ms.

---

Test

e2e/npm_translate_lock_replace_packages

All tests were cache hits

4 tests (100.0%) were fully cached saving 384ms.

---

Test

e2e/npm_translate_lock_subdir_patch

All tests were cache hits

1 test (100.0%) was fully cached saving 95ms.

---

Test

e2e/npm_translate_package_lock

All tests were cache hits

1 test (100.0%) was fully cached saving 31ms.

---

Test

e2e/npm_translate_yarn_lock

All tests were cache hits

1 test (100.0%) was fully cached saving 31ms.

---

Test

e2e/package_json_module

All tests were cache hits

1 test (100.0%) was fully cached saving 348ms.

---

Test

e2e/patch_from_repo

All tests were cache hits

1 test (100.0%) was fully cached saving 31ms.

---

Test

e2e/pnpm_lockfiles

All tests were cache hits

67 tests (100.0%) were fully cached saving 9s.

---

Test

e2e/pnpm_repo_install

All tests were cache hits

1 test (100.0%) was fully cached saving 2s.

---

Test

e2e/pnpm_workspace

All tests were cache hits

15 tests (100.0%) were fully cached saving 3s.

---

Test

e2e/pnpm_workspace_deps

All tests were cache hits

3 tests (100.0%) were fully cached saving 901ms.

---

Test

e2e/pnpm_workspace_rerooted

All tests were cache hits

15 tests (100.0%) were fully cached saving 5s.

---

Test

e2e/repo_mapping

All tests were cache hits

3 tests (100.0%) were fully cached saving 454ms.

---

Test

e2e/rules_foo

All tests were cache hits

2 tests (100.0%) were fully cached saving 258ms.

---

Test

e2e/runfiles

All tests were cache hits

1 test (100.0%) was fully cached saving 118ms.

---

Test

e2e/stamped_package_json

All tests were cache hits

1 test (100.0%) was fully cached saving 44ms.

---

Test

e2e/vendored_node

All tests were cache hits

1 test (100.0%) was fully cached saving 78ms.

---

Test

e2e/vendored_tarfile

All tests were cache hits

1 test (100.0%) was fully cached saving 31ms.

---

Test

e2e/verify_patches

All tests were cache hits

2 tests (100.0%) were fully cached saving 109ms.

---

Test

e2e/worker

All tests were cache hits

1 test (100.0%) was fully cached saving 35ms.

---

Test

e2e/workspace

All tests were cache hits

1 test (100.0%) was fully cached saving 35ms.

---

Buildifier      Format

[jbedard]

Maybe this is just re-inventing the CAP message from the original proposal?

@thesayyn were you thinking that a socket has a single subscription, and the CAP message is essentially declaring the options for that socket/subscription? This PR does similar but "subscribe" seems to imply that you can have more then one subscription per socket - which is probably more complicated and not worth it?

[thesayyn]

my worry with SUBSCRIBE is that it blocks cycle messages from being sent until the implementor tells us what it needs. the name implies that it can be sent N many times and every subscription needs an id to distinguish the CYCLE events from each other.

[jbedard]

I'm starting to think we should keep the existing behaviour where 1 socket = 1 subscription, and this PR (and the associated protocol spec changes) should instead implement the original CAP message - which I think aligns with your original idea?

[thesayyn]

Yes, and revisit the scope idea where as part of the CAP, we can decide what the user is asking for.