Fix part of #4064: Learner Analytics proto structures #4116
Conversation
This file contains 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
…lytics-proto-impl
Sarthak2601
changed the title
Sarthak2601
changed the title
|
Thanks @Sarthak2601. I'll need to follow up with a review on this tomorrow. |
BenHenning
requested changes
Jan 22, 2022
There was a problem hiding this comment.
Thanks @Sarthak2601! Took a first pass. I had a bunch of thoughts around the documentation context--PTAL.
utility/src/main/java/org/oppia/android/util/logging/EventBundleCreator.kt
Outdated
Show resolved
Hide resolved
Base automatically changed from
learner-analytics-platform-parameters
to
develop
January 25, 2022 04:54
BenHenning
approved these changes
Feb 10, 2022
|
Though FYI it appears there are broken CI checks--please fix and if there are any major changes that need re-review please readd me. |
|
Hi @Sarthak2601, I'm going to mark this PR as stale because it hasn't had any updates for 7 days. If no further activity occurs within 7 days, it will be automatically closed so that others can take up the issue. |
oppiabot
bot
added
the
stale
|
@vinitamurthi PTAL :) |
|
Was this closed accidentally? |
oppiabot
bot
removed
the
stale
|
I have reopened this PR , there are some conflicts but otherwise LGTM! |
vinitamurthi
approved these changes
Mar 11, 2022
|
Unassigning @vinitamurthi since they have already approved the PR. |
|
Hi @Sarthak2601, this PR is ready to be merged. Please address any remaining comments prior to merging, and feel free to merge this PR once the CI checks pass and you're happy with it. Thanks! |
# Conflicts: # app/src/main/res/layout/view_event_logs_event_log_item_view.xml
BenHenning
changed the title
6 tasks
BenHenning
added a commit
that referenced
this pull request
May 5, 2022
) ## Explanation <!-- - Explain what your PR does. If this PR fixes an existing bug, please include - "Fixes #bugnum:" in the explanation so that GitHub can auto-close the issue - when this PR is merged. --> Fix #4249 Fix part of #4064 Introduces the domain utilities necessary for logging learner analytics, but doesn't make them available for actual usage yet (that is being done in #4269 to keep this PR smaller & more focused). Some notes on the history of this PR: - This PR is a rebase of #4253 to remove dependencies on #2173 (which has since been merged into developer) for a much cleaner history. - This PR is pulling out elements from #4118, #4247, and #4248 which contained completed work by @Sarthak2601. - This PR extracts just the 'domain' pieces from the above, and changes a bunch about its architecture & adds tests. This PR should have little to no impact on the behavior of the app since the new logging functionality isn't being used yet, or is even accessible to broad components in the app. - This is starting as 'pt4' since it's continuing the work introduced in: #4114, #4115, and #4116. This PR makes a number of changes over the original design document and implementation, the most noteworthy being: - This PR organizes learner analytics logging into its own logger (and makes changes to event bundle creation & the generic ``OppiaLogger``). I think that we should move toward this pattern generally in the future rather than continuing with a generic ``OppiaLogger`` as it seems to help keep things much more focused. Existing logging should not be affected. - The notion of a device ID has been dropped as there's no reliable way retrieve such an ID (see https://developer.android.com/training/articles/user-data-ids). Instead, we're using a per-device ID (by leveraging ``PersistentCacheStore``), and have confirmed with study partners that this is workable. - The logging logic for the new logs was rearranged such that all new analytics logs will be logged for everyone, but the user and installation-tied IDs won't be logged in such cases (since they are more sensitive data). These events are generally useful for the platform, so we shouldn't restrict them as such. - Learner ID generation for profiles only occurs if the experiment is enabled, and otherwise stays empty. We may add future cleanup code to ensure it's erased across studies, but this at least lays the initial groundwork to keep such IDs separate when they aren't needed. For a high-level on the design, please refer to [this design document](https://docs.google.com/document/d/1c8lpH-IUvoU1t4LUoYNqNilP2e9yCnzGnSSG0yBxBrY/edit). Other noteworthy design choices: - ``DebugEventLogger`` was updated to call through to the real logger (as it makes event verification simpler in developer builds; normally analytics is off so this won't have any effects for the broader team) - Both ``DebugEventLogger`` and ``FakeEventLogger`` were updated to be thread-safe - Some extended functionality was added to ``FakeEventLogger`` - ``LearnerAnalyticsLogger`` is designed a bit differently compared to other domain classes in that it actually provides session-specific objects to the application-wide singleton graph (which is needed for logging certain situations, such as the user playing/stopping audio during a play session) - ``LoggingIdentifierController`` makes use of a lazy retrieval for session ID now (which is fine because it's guaranteed to compute exactly one initial ID) - ``StateFlow`` is used for easier cross-thread communication, including to expose internal asynchronous state across domain components (the only way we had to do this before was ``Deferred``, and that can be clunky; the new approach is much cleaner) - An ``EventLogSubject`` was introduced to make testing event logs easier. It's used extensively in tests for this feature, but most existing use cases weren't migrated. #4272 is tracking adding tests for this subject (hence the test file exemption). - There were TODOs introduced on #4064 to provide explicit clarity to reviewers on what needs to be changed in later PRs (as there's some things being introduced before the final PR that aren't actually used yet to help break up the project) - Multiple test suites verify behaviors with and without the feature enabled to be very explicit about what behavior occurs when - ``EventBundleCreatorTest`` in particular has very strict tests to ensure that sensitive IDs are logged exactly when expected (initially, never since they aren't turned on in this PR; this is fixed in a later PR) - ``ExplorationDataController`` was updated to introduce new play entrypoints, but these aren't "interesting" yet as the underlying ``ExplorationProgressController`` changes are coming in a later PR. Further, testing coverage technically removes checking ``playExploration``, but it'll be removed (and it's technically tested through the other functions since they call through). - A new ``ClipboardManager`` was introduced with the specific design of not allowing the broad app access to clipboard information from other apps. Instead, it provides an interface to confirm whether the app's known clipboard has been kept. A regex content check was added to ensure developers never use the clipboard service directly and instead use this manager. - ``PersistentCacheStore`` was updated to include a new ``primeInMemoryAndDiskCacheAsync`` function which works more predictably for initialization than ``primeInMemoryCacheAsync`` (formerly ``primeCacheAsync``). In particular, ``primeInMemoryCacheAsync`` is better for ensuring that the cache will quickly be read once it needs to be (and, if it isn't, will default in the same way the cache store normally defaults). However, there are cases when the app wants to change the default values such that: (1) the normal default is never used, (2) the default has to be computed and isn't cheap, and (3) it should never compute that default again once saved on disk. ``primeInMemoryAndDiskCacheAsync`` makes these assurances which, in turn, makes the installation ID cache store even possible without potential race conditions or breaking Dagger's cheap-initialization best practice. Test exemptions: all exemptions are annotations or interfaces except ``EventBundleCreatorTest`` (which is explained above). ## Essential Checklist <!-- Please tick the relevant boxes by putting an "x" in them. --> - [x] The PR title and explanation each start with "Fix #bugnum: " (If this PR fixes part of an issue, prefix the title with "Fix part of #bugnum: ...".) - [x] Any changes to [scripts/assets](https://github.com/oppia/oppia-android/tree/develop/scripts/assets) files have their rationale included in the PR explanation. - [x] The PR follows the [style guide](https://github.com/oppia/oppia-android/wiki/Coding-style-guide). - [x] The PR does not contain any unnecessary code changes from Android Studio ([reference](https://github.com/oppia/oppia-android/wiki/Guidance-on-submitting-a-PR#undo-unnecessary-changes)). - [x] The PR is made from a branch that's **not** called "develop" and is up-to-date with "develop". - [x] The PR is **assigned** to the appropriate reviewers ([reference](https://github.com/oppia/oppia-android/wiki/Guidance-on-submitting-a-PR#clarification-regarding-assignees-and-reviewers-section)). ## For UI-specific PRs only N/A -- This PR doesn't make UI changes, and existing flows shouldn't be affected. Commits: * strings for learner analytics * platform parameter impl for learner analytics * nit * nit * event action enum update * addition of contexts * nit * controller level logging and contexts * nit * nit fixes. * nit fixes. * event bundle modifications * sync status, logging identifiers, profile update, lifecycle owner * ui impl: part 1 -- basic * admin control strings * strings correction * strings correction * device id correction * exhaustive when fix. * exhaustive when fix. * todo formatting * nits. * nits. * collapsed contexts, added spacing, added comments * event action removal + nits * tests + dev options event logs fixes post event action removal * nits * removal of method for event action formatted string * nits, null context changes. * nits * reserved fixes and help index fix * bazel imports * bazel build fixes * test fixes * nit * logging identifier controller, module + uuid wrapper, real impl * logging identifier controller tests, fake uuid, tests * sync status manager + fake * logging methods, test setup * profile management, tests * sync status update. * lifecycle observer * Post-merge fixes + Bazel support. * Lots of reorganizing & changes. New tests and documentation have also been added. More broadly, this changes the device ID computation, but actually breaks it so more work will be needed in subsequent commits. * Lint fixes. * Post-merge fix (proper merge of maven_install). * Lint fixes (includes post-merge cleanups). * Lots of stuff. Restructured a lot of the UI, addressed most failing static checks (except KDocs and lint which will be in the follow-up commit), added tests, fixed copying, and generally finished the UI. Sync status seems broken, and it's not yet clear whether events are actually being logged (I need to investigate this). Analytics are disabled in local testing, so that might also be the reason for logs being stuck in an uploading state. * Documentation + lint fixes. This also changes the contract of ClipboardController. * Finish remaining planned tests. * Move over changes from learner-analytics-proto. * Manually pull in changes from 3d6c716. Note that this is operating on a different base). * Post-merge fixes. These at least ensure that the app can build, but many tests will still fail (which is fine seeing as much of this code is going to be split up soon, anyway). Rebase version: app build is no longer guaranteed. * Lint fixes. * Undo all learner analytics changes. I'll be pulling in specific components in specific PRs to organize the changes across 4 PRs. Note that I took this approach to preserve the history from the earlier commits. Those changes will still be included in this PR chain, just a bit awkwardly (i.e. it'll look like I introduced them originally, but that distinction is lost during the squash-and-merge, anyway). * Manually pull in non-app module changes. A bunch of work is still needed to finish these, and I'm still trying to figure out whether I can de-couple the module changes to make reviewing a bit nicer. * Post-merge fixes. All tests verified as building & passing. * Add sync status for no connectivity case. * Remove unnecessary sync manager. * Copy over changes from #4263. These are the domain changes needed for finishing learner analytics support. Cleanup, documentation, and testing all still need to be completed. * Add domain changes for AudioPlayerController. These originate from #4263. * Add missing Javadoc from #4263. * Finish tests & documentation. This also renames 'device ID' to be 'installation ID' for more correctness. * Lint fixes. * Fix OS-specific issue in ClipboardController. Co-authored-by: Sarthak Agarwal <agarwal.sarthak262012@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Explanation
Fixes part of #4064
This PR modifies the existing proto structure in
oppia_loggerto accommodate new learner study analytics events. There are 15 new analytics events that are being added as of now. The data related to these events is divided into three parts: Generic Data (common across all events), Exploration Data (common across all exploration-related events) and unique parameters. This PR adds proto definitions for event-specific contexts and actions along with definitions for Generic and Exploration Data.For reference:
Essential Checklist
For UI-specific PRs only
If your PR includes UI-related changes, then: