test(dada-client): characterize the untested surface (LIVE-35224)

[LucasWerey]

✅ Checklist

• npx changeset was attached.
• Covered by automatic tests. This PR is the tests.
• Impact of the changes:
  - No runtime impact — no production code was changed. The diff is 11 new test files plus a changeset.
  - CI signal only: the src/dada-client suite goes from 10 suites / 96 tests to 21 suites / 245 tests.
  - No QA pass needed. Nothing to test in the apps.

📝 Description

Problem. 14 source files under libs/ledger-live-common/src/dada-client had no colocated test, and the one existing api test covered only buildAssetsQueryParams and the two fetchAll*ByCategory helpers — not the RTK Query endpoints, getChunkedAssetsData, convertApiAssets, or transformAssetsResponse. The package is about to be relocated into DDD packages under LIVE-35223, and a green CI run would not have proven that relocation behaviour-neutral. The riskiest code in the package is also the least typed: createCurrencyDataSelector walks RTK Query's internal cache layout (queries[*].data.pages[*][key][id]) on Record<string, unknown>, so no refactor there can ever produce a type error.

Solution. 149 characterization tests pinning current behaviour, with no production change. These become the contract for the rest of the migration epic: they must keep passing unmodified, and a test that needs editing to accommodate a later change signals broken behaviour rather than a wrong test.

File	Tests	Pins
state-manager/__tests__/apiTransforms.test.ts	30	convertApiAssets, transformAssetsResponse, getChunkedAssetsData
hooks/__tests__/useMarketByCurrencies.test.tsx	19	both-fields-required, float rounding
entities/__tests__/selectorUtils.test.ts	16	cache-shape walk, cross-query bleed, first-match-wins, falsy-as-absent
utils/__test__/mergeAssetsDataPages.test.ts	13	shallow merge, last-page-wins, no input mutation
hooks/__tests__/useInterestRatesByCurrencies.test.tsx	13	APY-type allowlist, zero-is-a-value, isEqual stability
hooks/__tests__/useStockAssetIds.test.ts	13	Set wrapping, shared empty-set identity
hooks/__tests__/useDrawerConfiguration.test.ts	13	flag gating, config precedence
hooks/__tests__/useAssetData.test.ts	11	param narrowing, isLoading ‖ isFetching
hooks/__tests__/useChunkedAssetsData.test.ts	11	skip separation, full ErrorInfo shape
entities/__tests__/interestRateSelectors.test.ts	5	reads interestRates only
entities/__tests__/marketSelectors.test.ts	5	reads markets only

Four invariants that break silently are now pinned executably, since none of them can produce a type error:

1. assetsDataApi.reducerPath is the literal string "assetsDataApi" — read by string in selectorUtils.ts:25 and preloaded by three Storybook files.
2. The cache selectors scan every cache entry regardless of query args, so a value cached by one query is served to callers of another. Characterized, not endorsed.
3. convertApiAssets is deliberately lenient: it silently drops tokens whose parent chain is unknown, and synthesises a currency (color: "#999999", explorerViews: []) for cryptos missing from the local registry rather than dropping them.
4. getChunkedAssetsData succeeds if any chunk resolves.

The contract already proved itself

This branch was rebased onto 59 new develop commits, two of which changed dada-client production code (perf(modular-drawer): forward network filters to dada, adding networkIds to GetAssetsDataParams and buildAssetsQueryParams). Every characterization test passed unmodified through that rebase.

That change did open one gap, closed by the second commit here: networkIds also flows through getChunkedAssetsData, where — unlike currencyIds, which is split across chunks — it must be repeated in full on every request. buildAssetsQueryParams is covered by that change's own tests, but the chunked path's forwarding was not. Exactly the kind of detail a relocation drops silently.

Notes for reviewers

• convertApiAssets, transformAssetsResponse and emptyAssetsData are module-private and this task must not touch production code, so they are driven through the getAssetData / getChunkedAssetsData endpoints with a real store and a stubbed fetch. That pins the observable contract rather than internals, which is what the migration has to preserve anyway.
• getAssetData goes through fetchBaseQuery, so fetch receives a Request; the chunked path calls fetch with a string. requestedUrl() normalises both.
• Bug surfaced, not fixed here: an asset with id: "" makes CryptoCurrencyIdSchema.parse throw and fails the entire query, not just that asset. Pinned as current behaviour in apiTransforms.test.ts; the follow-up validation ticket (LIVE-35233) must not make it worse.
• The useMarketByCurrencies rounding table was verified against the implementation rather than assumed. Rounding is Math.round(x * 100) / 100, so results follow float representation, not decimal intuition: -2.345 → -2.35, -1.005 → -1, and an exact .5 rounds toward +∞ (0.125 → 0.13 but -0.125 → -0.12).
• useDrawerConfiguration is tested here although it is not DADA code — it only reads a feature flag. It is misfiled in this directory and gets relocated later in the epic, so it needs the same protection.
• Known, deliberately unfixed: apiTransforms.test.ts leaves a worker to be force-exited (A worker process has failed to exit gracefully). Several endpoints set keepUnusedDataFor: 86400, which parks a 24-hour timer. Releasing the subscription via unsubscribe() was tried and was not sufficient; the real fix is fake timers or a resetApiState teardown. Left out of scope on purpose so this PR stays tests-only.
• Changeset is patch on @ledgerhq/live-common — internal-only change, no runtime behaviour affected.

❓ Context

• JIRA or GitHub link: LIVE-35224 (epic: LIVE-35223)
• Stack: this is the bottom of a stack; #20285 builds on it.

---

🧐 Checklist for the PR Reviewers

• The code aligns with the requirements described in the linked JIRA or GitHub issue.
• The PR description clearly documents the changes made and explains any technical trade-offs or design decisions.
• There are no undocumented trade-offs, technical debt, or maintainability issues.
• The PR has been tested thoroughly, and any potential edge cases have been considered and handled.
• Any new dependencies have been justified and documented.
• Performance considerations have been taken into account. (changes have been profiled or benchmarked if necessary)

[github-actions]

Web Tools Build Status

Build	Status	Deployment
Web Tools Build	✅ Deployed	https://web-tools-ojyt4hozo-ledger-hq-prd.vercel.app
Native Storybook Build	⏭️ Skipped	—
React Storybook Build	⏭️ Skipped	—

[Copilot]

Pull request overview

Adds a comprehensive characterization-test suite around libs/ledger-live-common/src/dada-client to pin current behavior ahead of the planned DDD relocation/migration, without modifying production runtime code.

Changes:

• Add characterization tests for selector cache-walk behavior, API transforms, and RTK Query endpoints under src/dada-client.
• Add hook-level tests for query-arg shaping, loading/error semantics, memoization/stability guarantees, and edge cases.
• Add a patch changeset for @ledgerhq/live-common reflecting the test-only addition.

Reviewed changes

Copilot reviewed 12 out of 12 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
libs/ledger-live-common/src/dada-client/utils/test/mergeAssetsDataPages.test.ts	Characterizes merging semantics for paged assets data (shallow merge, last-write wins, pagination/currenciesOrder behavior).
libs/ledger-live-common/src/dada-client/state-manager/tests/apiTransforms.test.ts	Exercises public RTK Query endpoints to characterize private transform/conversion behaviors and chunked fetching semantics.
libs/ledger-live-common/src/dada-client/hooks/tests/useStockAssetIds.test.ts	Pins query args + Set-wrapping/memoization behavior for stock asset IDs hook.
libs/ledger-live-common/src/dada-client/hooks/tests/useMarketByCurrencies.test.tsx	Pins selection/filtering + rounding behavior for market data hook and reference stability.
libs/ledger-live-common/src/dada-client/hooks/tests/useInterestRatesByCurrencies.test.tsx	Pins allowlist/type filtering, zero-is-value semantics, and reference stability for interest rate hook.
libs/ledger-live-common/src/dada-client/hooks/tests/useDrawerConfiguration.test.ts	Characterizes feature-flag-driven drawer configuration precedence/merging and callback stability.
libs/ledger-live-common/src/dada-client/hooks/tests/useChunkedAssetsData.test.ts	Pins param shaping (skip separation), loading-state semantics, and ErrorInfo parsing.
libs/ledger-live-common/src/dada-client/hooks/tests/useAssetData.test.ts	Pins param narrowing behavior and `isLoading
libs/ledger-live-common/src/dada-client/entities/tests/selectorUtils.test.ts	Characterizes RTK Query cache-shape walk semantics and cross-query “first match wins” behavior.
libs/ledger-live-common/src/dada-client/entities/tests/marketSelectors.test.ts	Pins selector reads-only-from-markets behavior and empty/partial handling.
libs/ledger-live-common/src/dada-client/entities/tests/interestRateSelectors.test.ts	Pins selector reads-only-from-interestRates behavior and empty/field-preservation semantics.
.changeset/lucky-mirrors-observe.md	Patch changeset noting added characterization tests in @ledgerhq/live-common.

[github-actions]

Rsdoctor Bundle Diff Analysis

Found 7 projects in monorepo, 7 projects with changes.

📊 Quick Summary

Project	Total Size	Change
desktop-main	2.3 MB	-
desktop-preloader	7.1 KB	-
desktop-renderer	80.6 MB	-
desktop-webviewDappPreloader	36.9 KB	-
desktop-webviewPreloader	200.0 B	-
desktop-workers	36.8 KB	-
mobile	261.3 MB	-

📋 Detailed Reports (Click to expand)

📁 desktop-main

Path: rsdoctor/desktop-main/rsdoctor-data.json

⚠️ No baseline data found - Unable to perform comparison analysis

Metric	Current	Baseline	Change
📊 Total Size	2.3 MB	-	-
📄 JavaScript	2.2 MB	-	-
🎨 CSS	0 B	-	-
🌐 HTML	0 B	-	-
📁 Other Assets	135.3 KB	-	-

📁 desktop-preloader

Path: rsdoctor/desktop-preloader/rsdoctor-data.json

⚠️ No baseline data found - Unable to perform comparison analysis

Metric	Current	Baseline	Change
📊 Total Size	7.1 KB	-	-
📄 JavaScript	5.3 KB	-	-
🎨 CSS	0 B	-	-
🌐 HTML	0 B	-	-
📁 Other Assets	1.8 KB	-	-

📁 desktop-renderer

Path: rsdoctor/desktop-renderer/rsdoctor-data.json

⚠️ No baseline data found - Unable to perform comparison analysis

Metric	Current	Baseline	Change
📊 Total Size	80.6 MB	-	-
📄 JavaScript	29.3 MB	-	-
🎨 CSS	183.1 KB	-	-
🌐 HTML	1.8 KB	-	-
📁 Other Assets	51.2 MB	-	-

📁 desktop-webviewDappPreloader

Path: rsdoctor/desktop-webviewDappPreloader/rsdoctor-data.json

⚠️ No baseline data found - Unable to perform comparison analysis

Metric	Current	Baseline	Change
📊 Total Size	36.9 KB	-	-
📄 JavaScript	36.9 KB	-	-
🎨 CSS	0 B	-	-
🌐 HTML	0 B	-	-
📁 Other Assets	0 B	-	-

📁 desktop-webviewPreloader

Path: rsdoctor/desktop-webviewPreloader/rsdoctor-data.json

⚠️ No baseline data found - Unable to perform comparison analysis

Metric	Current	Baseline	Change
📊 Total Size	200.0 B	-	-
📄 JavaScript	200.0 B	-	-
🎨 CSS	0 B	-	-
🌐 HTML	0 B	-	-
📁 Other Assets	0 B	-	-

📁 desktop-workers

Path: rsdoctor/desktop-workers/rsdoctor-data.json

⚠️ No baseline data found - Unable to perform comparison analysis

Metric	Current	Baseline	Change
📊 Total Size	36.8 KB	-	-
📄 JavaScript	36.8 KB	-	-
🎨 CSS	0 B	-	-
🌐 HTML	0 B	-	-
📁 Other Assets	0 B	-	-

📁 mobile

Path: rsdoctor/mobile/rsdoctor-data.json

⚠️ No baseline data found - Unable to perform comparison analysis

Metric	Current	Baseline	Change
📊 Total Size	261.3 MB	-	-
📄 JavaScript	110.4 MB	-	-
🎨 CSS	0 B	-	-
🌐 HTML	0 B	-	-
📁 Other Assets	150.9 MB	-	-

Generated by Rsdoctor GitHub Action

[Copilot]

Pull request overview

Copilot reviewed 12 out of 12 changed files in this pull request and generated no new comments.

Suppressed comments (3)

libs/ledger-live-common/src/dada-client/state-manager/tests/apiTransforms.test.ts:100

• Avoid jest.restoreAllMocks() here: it restores all spies/mocks (including those installed by global jest setup), which can break unrelated tests when running the suite in parallel. Prefer restoring only the local fetchSpy for this file.

afterEach(() => {
  jest.restoreAllMocks();
});

libs/ledger-live-common/src/dada-client/state-manager/tests/apiTransforms.test.ts:96

• jest.spyOn(globalThis, "fetch") will call the real undici fetch when a test forgets to stub a response, which can accidentally hit the network and make failures slow/flaky. It’s safer for this suite to default fetch to throwing unless a test explicitly mockResolvedValueOnce/mockRejectedValueOnce.

beforeEach(() => {
  fetchSpy = jest.spyOn(globalThis, "fetch");
});

libs/ledger-live-common/src/dada-client/hooks/tests/useDrawerConfiguration.test.ts:15

• This test mocks useFeature, but useFeature is a Redux selector hook (it uses useSelector). Repo testing guidance prefers exercising feature flags via a store state (e.g., a Provider + featureFlags.resolved.ptxEarnDrawerConfiguration) rather than mocking the hook, so the test better reflects real behavior and avoids mock drift.

import { useFeature } from "@features/platform-feature-flags";
import { useDrawerConfiguration } from "../useDrawerConfiguration";

jest.mock("@features/platform-feature-flags", () => ({ useFeature: jest.fn() }));

const mockUseFeature = jest.mocked(useFeature);

// eslint-disable-next-line @typescript-eslint/no-explicit-any
const flag = (value: unknown) => mockUseFeature.mockReturnValue(value as any);

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud