Core app review & feedback API (#3665)#5268
Open
shai-almog wants to merge 7 commits into
Open
Conversation
Introduces com.codename1.appreview.AppReview: requests the platform's native store-review prompt (SKStoreReviewController on iOS, Play In-App Review on Android) when available and falls back to a Codename One drawn rating widget elsewhere. Adds an opt-in engagement scheduler (launch / days-installed / cool-down thresholds persisted via Preferences) plus a manual requestReview() trigger, and a smart feedback split that routes high ratings to the store and low ratings to a pluggable FeedbackListener (e-mail helper provided). Plumbing follows the existing share()/dial() pattern: - CodenameOneImplementation.isNativeInAppReviewSupported() / requestNativeInAppReview(), exposed via Display and CN. - iOS: IOSNative.requestAppStoreReview() bridges to SKStoreReviewController guarded by a new CN1_USE_APPREVIEW macro (symbol always emitted, StoreKit body compiled only when used). - Android: AppReviewSupport drives the Play review flow via reflection so the port compiles without the extra dependency. Native dependencies are added implicitly (no build hint): the Android/iOS builders detect com.codename1.appreview usage during their class scan and only then inject com.google.android.play:review / link StoreKit.framework. Tests & docs: - maven/core-unittests: AppReviewTest covers the scheduler decision logic. - docs/developer-guide: new App-Review chapter. - scripts/hellocodenameone: AppReviewDialogScreenshotTest for the widget. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Contributor
|
Developer Guide build artifacts are available for download from this workflow run:
Developer Guide quality checks: |
Collaborator
Author
Android screenshot updatesCompared 130 screenshots: 129 matched, 1 missing reference.
Native Android coverage
Benchmark ResultsDetailed Performance Metrics
|
Contributor
Cloudflare Preview
|
Collaborator
Author
Native Linux port (x64)Compared 129 screenshots: 128 matched, 1 missing reference.
|
Collaborator
Author
Native Linux port (arm64)Compared 129 screenshots: 128 matched, 1 missing reference.
|
Collaborator
Author
Native Windows port (cross-compiled)Compared 127 screenshots: 126 matched, 1 missing reference.
Benchmark ResultsDetailed Performance Metrics
|
Collaborator
Author
JavaScript port screenshot updatesCompared 123 screenshots: 122 matched, 1 missing reference.
|
Collaborator
Author
Mac native screenshot updatesCompared 130 screenshots: 129 matched, 1 missing reference.
Benchmark Results
Detailed Performance Metrics
|
Collaborator
Author
|
Compared 211 screenshots: 211 matched. |
Collaborator
Author
iOS screenshot updatesCompared 126 screenshots: 125 matched, 1 missing reference.
Benchmark Results
Build and Run Timing
Detailed Performance Metrics
|
Collaborator
Author
iOS Metal screenshot updatesCompared 130 screenshots: 129 matched, 1 missing reference.
Benchmark Results
Build and Run Timing
Detailed Performance Metrics
|
- RatingDialog now presents as a bottom Sheet (swipe to dismiss) instead of a blocking modal Dialog, addressing the "too intrusive" feedback. - Stars use a single-row GridLayout so they no longer wrap on narrow displays (the Android screenshot showed the flow-layout wrap). - Fix SpotBugs LI_LAZY_INIT_STATIC by eager-initializing the AppReview singleton. - Rewrite the developer-guide chapter to satisfy the Vale prose gate (contractions, adverbs, quote punctuation) and describe the Sheet UI. - Screenshot test mirrors the single-row star grid. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Make RatingDialog final (ClassWithOnlyPrivateConstructorsShouldBeFinal). - Simplify shouldPrompt() boolean return (SimplifyBooleanReturns). - Add @OverRide to anonymous Runnable/SuccessCallback/ActionListener implementations (MissingOverride). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Contributor
✅ Continuous Quality ReportTest & Coverage
Static Analysis
Generated automatically by the PR CI workflow. |
Seeds scripts/ios/screenshots-watch/AppReviewDialog.png from the watch suite's streamed render of the rating sheet. The watch runner gates on missing_expected (CN1SS_ALLOWED_MISSING=0); the other platforms treat a new golden as non-gating, and build-ios-metal is intentionally left without a baseline since that job is flaky. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The test previously rendered an inline Dialog-styled container; it now builds the same content as RatingDialog and calls Sheet.show() (following the SheetScreenshotTest pattern) so the capture shows the actual sheet chrome. The native store review widgets are OS-drawn, throttled overlays outside Display.screenshot(), so they can't be captured by the harness; this covers the Codename One fallback sheet that renders elsewhere. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The committed watch golden was the old inline-Dialog render; the test now shows the real Sheet, so the golden must be regenerated. Removing it both avoids a false "different" and (being a screenshots-watch change) re-runs the full screenshot suite against the new test so the fresh render can be captured. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Regenerate scripts/ios/screenshots-watch/AppReviewDialog.png from the real-Sheet render of the updated test. - Developer guide: embed the fallback rating Sheet screenshot (a cropped desktop render showing the bottom sheet with the single-row star strip) and describe the native iOS/Android prompts in prose, since those OS-drawn widgets are outside Display.screenshot() and can't be captured. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
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.
Resolves #3665.
Adds a core app-review & feedback API (
com.codename1.appreview) instead of a cn1lib. It uses the platform's native store-review prompt when available and falls back to a built-in Codename One widget everywhere else, while giving the app full control over when feedback is requested.What's included
Public API —
com.codename1.appreviewAppReview— singleton facade with two usage styles:AppReview.getInstance().requestReview()— ask whenever it makes sense.minimumLaunches/minimumDaysInstalled/daysBetweenPrompts) and callregisterSession()each launch; state is persisted inPreferencesand the user is never re-prompted after rating or opting out.highRatingThreshold) to the store and low ratings to a pluggableFeedbackListener(e-mail helper provided), so unhappy users reach you privately instead of leaving a 1-star public review.RatingDialog— the built-in fallback widget (stars + feedback), package-private.Native plumbing (mirrors
share()/dial())CodenameOneImplementation.isNativeInAppReviewSupported()/requestNativeInAppReview(SuccessCallback<Boolean>), exposed viaDisplayandCN.IOSNative.requestAppStoreReview()→SKStoreReviewController.requestReview(iOS ≥ 10.3), guarded by a newCN1_USE_APPREVIEWmacro (symbol always emitted so ParparVM links; StoreKit body compiled only when used).AppReviewSupportdrives the Play In-App Review flow via reflection, so the port compiles without the extra dependency;isNativeInAppReviewSupported()returns false when the lib wasn't bundled.Implicit dependency inclusion (no build hint)
The Android/iOS builders already scan the app's classes to auto-enable native deps. This PR extends that scan so referencing
com.codename1.appreview(or theCN/Displayreview methods) automatically injectscom.google.android.play:reviewon Android and linksStoreKit.framework+ flipsCN1_USE_APPREVIEWon iOS. Apps that never ask for a review carry no extra weight.Tests & docs
maven/core-unittests/.../appreview/AppReviewTest.java— 8 JUnit5 tests covering the scheduler decision logic (run:mvn -DunitTests -pl core-unittests test -Dtest=AppReviewTest, all green).docs/developer-guide/App-Review.asciidoc, registered in the master document.AppReviewDialogScreenshotTestinscripts/hellocodenameone, registered inCn1ssDeviceRunner. It is a brand-new golden, so each platform's screenshot job reports it asmissing_expected(non-gating) until the baselines are recorded via the usual flow.Verified
mvn compileof the maven-plugin module (both builders) → BUILD SUCCESS.AppReviewSupportcompiles againstandroid.jar; the screenshot test compiles against core.Note for maintainers
The same implicit-detection edits were applied to the out-of-repo cloud build server copies (
~/dev/CodenameOne/BuildDaemon,-analytics,-winpkg) since those aren't part of this repo — they need to be deployed separately for the native dep to be injected on cloud builds.🤖 Generated with Claude Code