Depends on D15253
Create array of test vectors for parseChronikTx unit tests. Mocks are updated to have tx and parsed keys, and renamed to better reflect test case
Details
Details
- Reviewers
Fabien - Group Reviewers
Restricted Project - Commits
- rABC7c3c410622a7: [Cashtab] Refactor unit tests for parseChronikTx
npm test
Diff Detail
Diff Detail
- Repository
- rABC Bitcoin ABC
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
| cashtab/src/chronik/fixtures/vectors.js | ||
|---|---|---|
| 23 ↗ | (On Diff #44501) | What's the point of having the inputs in one file and the expected output in another ? This is just making it more difficult to read/change/add a test. Imo you should have the inputs and expected output in the same block, so it's super clear what the inputs you're crafting are expected to produce. |
| cashtab/src/chronik/fixtures/vectors.js | ||
|---|---|---|
| 23 ↗ | (On Diff #44501) | In this case I think the test vectors are easier to read and maintain using variables at these keys, because the actual tx objects are quite large and cumbersome. This is why they live in mocks.js -- standard purpose of this file in Cashtab is to abstract away the storage of large / complicated mocks that are for the most part copy pasted chronik output. To add a test here, we
We could just keep the mocks here, but I think having large objects inside the vector keys makes them harder to read / understand.
In mocks.js, each tx object includes the input at key tx and the output at key parsed. I could move these to live in this array, but then it becomes much harder to see what the vectors are because tx is a big object. |