test(e2e): add setup for E2E testing using bublik-e2e tool - #71
Draft
okt-limonikas wants to merge 12 commits into
Draft
test(e2e): add setup for E2E testing using bublik-e2e tool#71okt-limonikas wants to merge 12 commits into
bublik-e2e tool#71okt-limonikas wants to merge 12 commits into
Conversation
okt-limonikas
force-pushed
the
e2e-testing
branch
3 times, most recently
from
June 26, 2026 13:25
0f33e8b to
132a2ca
Compare
Signed-off-by: Danil Kostromin <danil.kostromin@icloud.com>
okt-limonikas
force-pushed
the
e2e-testing
branch
from
August 12, 2026 13:25
132a2ca to
9107c67
Compare
Wire pnpm run e2e:features:check into the pre-run gate so a scenario without a Playwright test fails locally and in CI (which invokes task e2e:types:check before building the stack).
`task --list` offered 18 e2e tasks, most of them variations on the same three actions (up:reuse, up:fresh, run, run:fresh, run:reuse, down, down:fresh, reset, clean), so picking one meant reading the Taskfile. Five are listed now — reset, up, down, test, test:ui. The rest stay runnable but drop their `desc`, so they show up under `task --list-all` as the debugging helpers they are; the steps that are meaningless or dangerous on their own are `internal: true`. `e2e:up` seeds only when the instance does not already have the runs, so it is cheap to re-run. The probe checks the manifest *and* asks the API whether one of its run ids still exists: a database wiped underneath a stale manifest must not skip a seed the stack needs. `e2e:reset` now also clears Playwright's reports, traces and cached auth state. Nothing did, so a "clean" run could still be reading a previous run's output. Campaign parsing moves to the bublik-e2e CLI (`--plan`), which already owned every validator scripts/e2e_plan.py duplicated. What is left is this repo's concern: the path guard that keeps a cleanup inside the directories E2E may write to, the Compose-project guard, and the seed probe. The plan itself becomes YAML, so each run group is one line and the days can carry comments; `task e2e:plan` expands it. The 50-line nested readiness probe becomes scripts/e2e_ready.sh, a flat loop over the endpoints a test can touch. E2E tasks live in Taskfile.e2e.yml. Their environment is shared through a YAML anchor rather than the included file's root `env:`, which Task merges into the whole run — it would have followed the production stack into `task docker:up` and pointed it at the E2E data directory.
`task e2e:report:bundle` converts the last Playwright JSON report into a Bublik
bundle and publishes it, printing the import URL; `-- --import` imports it too.
It is a helper, not wired into e2e:test or e2e:ci, so a suite run stays a suite
run.
Results are published to a sibling of the fixture directory, never inside it.
`generate` replaces its entire publish directory on every seed and
`scripts/e2e.py clean` deletes it on reset, while accumulating result runs
across resets is the whole point of importing them. `e2e-results` sits outside
that blast radius — `_resolve_allowed` confines the deletion to
{DATA_DIR}/logs/logs/e2e — and stays under the served logs root, so the
instance's existing LOGS_BASES covers its URL.
Also bumps bublik-ui for the json reporter the task reads.
Claude-Session: https://claude.ai/code/session_01CoxryDN5irXCVFgMrBWEtz
Taskfile.e2e.yml existed to keep the E2E Compose project and data directory from leaking into the production tasks, but the per-task env anchor it already used does that just as well inside one file — and the include cost API_URL forwarded by hand plus a ":docker:setup" escape on every cross-reference. Keep six tasks instead of seventeen: up, seed, test and down are the loop; logs and types:check stay runnable but unlisted. The rest were one-line wrappers around bublik-e2e subcommands, so call those directly. Splitting up from seed also gives CI somewhere to put the chmod that lets the host seeder write into the containers' data dir. Vars gain an E2E_ prefix now that they share a namespace with the production ones; the env var names they feed are unchanged. Claude-Session: https://claude.ai/code/session_01KDfUVbrKGpHugNSefSoUmT
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.