tom/HealthLog
1
0 Fork 0
mirror of https://github.com/MBombeck/HealthLog.git synced 2026-11-09 13:01:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

e2e: pin the auto-read contract per document read state

Browse Source 
The auto-read spec pinned the old collapse rule: with auto-read ON the
sheet always dropped the AI block. Since the read action stays visible
for a document that has no content index yet, the collapse assertion
now runs against the indexed content fixture, and a second test pins
the recovery path: an unread document keeps its AI section and manual
read under auto-read. Refs #777.
This commit is contained in:
Marc-André Bombeck committed 2026-08-13 07:35:15 +02:00
1 parent ff88dd1f8f
commit 91d4908204
1 file changed
+35 -11
Download Patch File Download Diff File Expand all files Collapse all files
e2e/documents-auto-read.spec.ts
+35 -11
View File
@@ -23,6 +23,7 @@ import {
ensureVaultFixture,
ensureVaultAiFixture,
AI_PROBE_DOC_ID,
CONTENT_DOC_ID,
} from "./setup/vault-fixture";
async function fulfilJson(
@@ -249,24 +250,23 @@ test.describe("automatic AI reading — vault per-document contract", () => {
await expect(sheet.locator('[data-slot="document-read-ai"]')).toBeVisible();
});
test("with auto-read ON the sheet drops the manual AI action row", async ({
test("with auto-read ON the sheet drops the AI row only for a read document", async ({
page,
}) => {
// Auto-read reads every upload with no per-document tap, so the manual
// Read / Suggest actions are redundant and the WHOLE block collapses
// (v1.28.22 — no status chrome; provenance lives on the vault card).
// Auto-read reads every upload with no per-document tap, so once a
// document IS read the manual Read / Suggest actions are redundant and
// the WHOLE block collapses (v1.28.22 — no status chrome; provenance
// lives on the vault card). `CONTENT_DOC_ID` ships with a content index,
// so it is the read-document case.
await mockDocumentCapability(page, { egress: "external", autoRead: true });
await page.goto(`/documents?doc=${AI_PROBE_DOC_ID}`);
await page.goto(`/documents?doc=${CONTENT_DOC_ID}`);
const sheet = page.getByRole("dialog");
await expect(sheet).toBeVisible();
// v1.28.22 — with auto-read ON and nothing pending the WHOLE block
// collapses (no status chrome; provenance lives on the vault card), so the
// sheet content starts straight with the document fields. The footer's
// download action is the loaded-state anchor (it renders only once the
// document GET resolved), so the absence checks below cannot pass
// vacuously against the loading skeleton.
// The footer's download action is the loaded-state anchor (it renders
// only once the document GET resolved), so the absence checks below
// cannot pass vacuously against the loading skeleton.
await expect(
sheet.locator('[data-slot="document-download"]'),
).toBeVisible();
@@ -278,4 +278,28 @@ test.describe("automatic AI reading — vault per-document contract", () => {
);
await expect(sheet.locator('[data-slot="assist-suggest"]')).toHaveCount(0);
});
test("with auto-read ON an unread document keeps its read action", async ({
page,
}) => {
// v1.37.13 — auto-read reads new uploads, but a document that has no
// content index yet (upload predates the switch, or its read failed)
// still needs a way in: the AI section stays, with the manual read as
// the per-document recovery path. Collapsing it here would strand the
// document with no route to a read at all. The probe document is seeded
// without a content index, so it is exactly that case.
await mockDocumentCapability(page, { egress: "external", autoRead: true });
await page.goto(`/documents?doc=${AI_PROBE_DOC_ID}`);
const sheet = page.getByRole("dialog");
await expect(sheet).toBeVisible();
await expect(
sheet.locator('[data-slot="document-download"]'),
).toBeVisible();
await expect(
sheet.locator('[data-slot="document-ai-section"]'),
).toHaveCount(1);
await expect(sheet.locator('[data-slot="document-read-ai"]')).toBeVisible();
});
});