-
Notifications
You must be signed in to change notification settings - Fork 435
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix issue #604 #793
Fix issue #604 #793
Conversation
Cc @seanpdoyle |
The issues outlined by [hotwired#793][] are resolved by the introduction of the `FrameVisit` object and its lifecycle. To ensure that behavior is fixed, this commit cherry-picks the test coverage introduced in that branch. [hotwired#793]: hotwired#793
@manuelpuyol I've cherry-picked the test coverage you've introduced in this PR to #430, and the test suite passes. That pull request introduces the concept of a Having said that, I've applied the following diff to diff --git a/src/tests/functional/frame_navigation_tests.ts b/src/tests/functional/frame_navigation_tests.ts
index 3a2fadd..ac6eb03 100644
--- a/src/tests/functional/frame_navigation_tests.ts
+++ b/src/tests/functional/frame_navigation_tests.ts
@@ -1,5 +1,12 @@
import { test } from "@playwright/test"
-import { getFromLocalStorage, nextEventNamed, nextEventOnTarget, pathname, scrollToSelector } from "../helpers/page"
+import {
+ getFromLocalStorage,
+ nextBeat,
+ nextEventNamed,
+ nextEventOnTarget,
+ pathname,
+ scrollToSelector,
+} from "../helpers/page"
import { assert } from "chai"
test("test frame navigation with descendant link", async ({ page }) => {
@@ -97,3 +104,38 @@ test("test promoted frame navigations are cached", async ({ page }) => {
assert.equal(await page.getAttribute("#tab-frame", "src"), null, "caches one.html without #tab-frame[src]")
assert.equal(await page.getAttribute("#tab-frame", "complete"), null, "caches one.html without [complete]")
})
+
+test("test canceling frame requests don't mutate the history", async ({ page }) => {
+ await page.goto("/src/tests/fixtures/tabs.html")
+
+ await page.click("#tab-2")
+
+ await nextEventOnTarget(page, "tab-frame", "turbo:frame-load")
+ await nextEventNamed(page, "turbo:load")
+
+ assert.equal(await page.textContent("#tab-content"), "Two")
+ assert.equal(pathname((await page.getAttribute("#tab-frame", "src")) || ""), "/src/tests/fixtures/tabs/two.html")
+ assert.equal(await page.getAttribute("#tab-frame", "complete"), "", "sets [complete]")
+
+ // This request will be canceled
+ page.click("#tab-1")
+ await page.click("#tab-3")
+
+ await nextEventOnTarget(page, "tab-frame", "turbo:frame-load")
+ await nextEventNamed(page, "turbo:load")
+
+ assert.equal(await page.textContent("#tab-content"), "Three")
+ assert.equal(pathname((await page.getAttribute("#tab-frame", "src")) || ""), "/src/tests/fixtures/tabs/three.html")
+
+ await page.goBack()
+ await nextEventNamed(page, "turbo:load")
+
+ assert.equal(await page.textContent("#tab-content"), "Two")
+ assert.equal(pathname((await page.getAttribute("#tab-frame", "src")) || ""), "/src/tests/fixtures/tabs/two.html")
+
+ // Make sure the frame is not mutated after some time.
+ await nextBeat()
+
+ assert.equal(await page.textContent("#tab-content"), "Two")
+ assert.equal(pathname((await page.getAttribute("#tab-frame", "src")) || ""), "/src/tests/fixtures/tabs/two.html")
+}) Were you able to force the tests to fail? |
The issues outlined by [hotwired#793][] are resolved by the introduction of the `FrameVisit` object and its lifecycle. To ensure that behavior is fixed, this commit cherry-picks the test coverage introduced in that branch. [hotwired#793]: hotwired#793
@seanpdoyle hmm maybe the test is not ideal, since I still can see the bug in main Screen.Recording.2022-11-30.at.12.59.42.PM.mov |
The issues outlined by [hotwired#793][] are resolved by the introduction of the `FrameVisit` object and its lifecycle. To ensure that behavior is fixed, this commit cherry-picks the test coverage introduced in that branch. [hotwired#793]: hotwired#793
The issues outlined by [hotwired#793][] are resolved by the introduction of the `FrameVisit` object and its lifecycle. To ensure that behavior is fixed, this commit cherry-picks the test coverage introduced in that branch. [hotwired#793]: hotwired#793
Closes #604
When quickly clicking between frames, even though the fetch request is canceled, the page's HTML is mutated, setting a new
src
to the frame element. This change will cause thepageSnapshot
to be incorrect, so when we navigate back, theturbo-frame
will have an incorrectsrc
causing a new fetch and the bug shown in the issue.This PR updates frames to know what was their
previousSrc
so they can revert back in case the request is canceled.