[aform, atable] update test coverage for Stonecrop packages

[Alchez]

Closes #108.

---

Todo:

• ATable components
  • AExpansionRow
  • ARow

[Alchez]

@agritheory This PR switches the Vitest reporter from v8 to istanbul.

For building the coverage report, Vitest defaults to the v8 reporter. Unfortunately, v8's AST considers things like style tags inside Vue files as something that requires testing as well, which means our coverage reports will always have some in-built inaccuracies.

On the other hand, the istanbul reporter has much finer control over what counts as coverage, and it seems to report general coverage better as well. Additionally, the docs over at v8 says that they're trying to integrate the reporter's coverage information into istanbul as well.

[github-actions]

Coverage Report for ./atable

Status	Category	Percentage	Covered / Total
🟢	Lines	73.44% (🎯 70%)	130 / 177
🟢	Statements	73.07% (🎯 70%)	133 / 182
🟢	Functions	78.12% (🎯 70%)	25 / 32
🔴	Branches	47.79% (🎯 70%)	65 / 136

File Coverage

File	Stmts	% Branch	% Funcs	% Lines	Uncovered Lines
Changed Files
atable/src/components/ACell.vue	94.64%	56.75%	100%	94.64%	64, 104-107, 158
atable/src/components/ARow.vue	55.17%	47.82%	60%	55.17%	49, 54, 60-68, 61-65, 62, 64, 67, 80, 84, 86-91, 87-90, 93-98, 12
atable/src/components/ATable.vue	88.46%	50%	100%	91.66%	156-158, 157, 157
Unchanged Files
atable/src/components/AExpansionRow.vue	0%	0%	0%	0%	23, 36-38, 40-42, 41, 44-63, 45-51, 47-49, 53-55, 54, 57-62, 1, 3
atable/src/components/ATableHeader.vue	100%	50%	100%	100%
atable/src/components/ATableModal.vue	75%	100%	0%	75%	21
atable/src/components/index.ts	70.45%	50%	88.88%	71.42%	47-49, 48, 61, 104-116, 105-113, 106-112, 108-111, 110, 114-116, 115

Generated in workflow #225

[github-actions]

Coverage Report for ./aform

Status	Category	Percentage	Covered / Total
🟢	Lines	90.35% (🎯 70%)	206 / 228
🟢	Statements	90.47% (🎯 70%)	209 / 231
🟢	Functions	87.5% (🎯 70%)	42 / 48
🟢	Branches	74.44% (🎯 70%)	67 / 90

File Coverage

File	Stmts	% Branch	% Funcs	% Lines	Uncovered Lines
Changed Files
aform/src/components/form/ADate.vue	85.71%	50%	100%	85.71%	42
aform/src/components/form/ADatePicker.vue	96.61%	90%	92.85%	96.49%	65, 27
aform/src/components/form/ADropdown.vue	91.83%	76.92%	84.61%	91.83%	55, 84-85, 93
aform/src/components/form/AFieldset.vue	90%	100%	50%	90%	8
aform/src/components/form/ATextInput.vue	100%	100%	100%	100%
aform/src/components/utilities/Login.vue	88.88%	81.81%	100%	88.88%	82-84
aform/src/directives/mask.ts	84.09%	59.09%	100%	83.72%	32-33, 40, 49, 64, 97, 102
Unchanged Files
aform/src/components/AForm.vue	90%	40%	85.71%	90%	41-43, 42
aform/src/components/base/CollapseButton.vue	100%	100%	100%	100%
aform/src/components/form/ACheckbox.vue	100%	100%	100%	100%
aform/src/components/form/AComboBox.vue	0%	100%	0%	0%	2-1, 7-3
aform/src/components/form/ANumericInput.vue	100%	100%	100%	100%

Generated in workflow #225

[Alchez]

@agritheory, the aform application code was pretty useful to test with Vitest since it's all isolated components with unique functions that can be tested headlessly.

However, I'm writing atable tests and I'm wondering if those are better suited to be written in a non-headless, browser framework Cypress or WebdriverIO. This is mostly because the components are coupled and there's a lot of visual assertions to be possibly made on the components based on the usage options (and Istanbul/v8 are erratic with coverage on child components).

What do you think?