captoys/diigo-cache/tests/fill.test.ts

import { env } from "cloudflare:workers";
import { evictDurableObject, runDurableObjectAlarm, runInDurableObject } from "cloudflare:test";
import { describe, expect, it, vi } from "vitest";
import worker from "../src/index.js";
import { toWebkey } from "../src/webkeys.js";

// goal 1e (capsite checklist): a job that grabs every page. Fake
// accounts: tests/fake-webapp.js ("fake-webapp-id": 30 items = 2 pages
// of 24, every 3rd private).

type TestEnv = {
	CACHE_DO: DurableObjectNamespace<import("../src/cache.js").DiigoCache>;
	PUBLIC_DO: DurableObjectNamespace<import("../src/cache.js").PublicCache>;
	WEBAPP: { calls(): Promise<number> };
};
const testEnv = env as unknown as TestEnv;

const wired = async (webappId = "fake-webapp-id") => {
	const stub = testEnv.CACHE_DO.get(testEnv.CACHE_DO.newUniqueId());
	await stub.setup({ webappId });
	return stub;
};

/** run alarms until none is scheduled (bounded, so a runaway job fails) */
const drain = async (stub: DurableObjectStub, max = 20) => {
	let ran = 0;
	while (ran < max && (await runDurableObjectAlarm(stub))) ran += 1;
	expect(ran).toBeLessThan(max);
	return ran;
};

describe("fill job (goal 1e)", () => {
	it("fills every page; afterwards every page is a hit with no webapp call", async () => {
		const stub = await wired();
		const atStart = await testEnv.WEBAPP.calls();
		expect((await stub.startFill()).state).toBe("running");
		await drain(stub);
		const status = await stub.fillStatus();
		expect(status).toMatchObject({ state: "done", pages: 2, fetched: 2 });
		// one diigo call per page fetched (no extra page-1 fetch per miss)
		expect((await testEnv.WEBAPP.calls()) - atStart).toBe(2);

		const before = await testEnv.WEBAPP.calls();
		for (const n of [1, 2]) expect((await stub.page(n, 24)).cached).toBe(true);
		expect(await testEnv.WEBAPP.calls()).toBe(before);
	});

	it("advances pagesPerStep pages per alarm, with progress readable between steps", async () => {
		const stub = await wired();
		await stub.startFill({ pagesPerStep: 1 });
		expect(await runDurableObjectAlarm(stub)).toBe(true);
		expect(await stub.fillStatus()).toMatchObject({ state: "running", nextPage: 2, pages: 2 });
		await drain(stub);
		expect((await stub.fillStatus())?.state).toBe("done");
	});

	it("survives eviction: progress is in storage, not memory", async () => {
		const stub = await wired();
		await stub.startFill({ pagesPerStep: 1 });
		await runDurableObjectAlarm(stub);
		await evictDurableObject(stub);
		expect(await stub.fillStatus()).toMatchObject({ state: "running", nextPage: 2 });
		await drain(stub);
		expect((await stub.fillStatus())?.state).toBe("done");
	});

	it("starting again while running rejoins the same job", async () => {
		const stub = await wired();
		const first = await stub.startFill({ pagesPerStep: 1 });
		await runDurableObjectAlarm(stub);
		const again = await stub.startFill({ pagesPerStep: 1 });
		expect(again.startedAt).toBe(first.startedAt);
		expect(again.nextPage).toBe(2);
	});

	it("skips pages already cached", async () => {
		const stub = await wired();
		await stub.page(1, 24);
		await stub.startFill();
		await drain(stub);
		expect(await stub.fillStatus()).toMatchObject({ state: "done", fetched: 1 });
	});

	it("stops when asked; cached pages stay", async () => {
		const stub = await wired();
		await stub.startFill({ pagesPerStep: 1 });
		await runDurableObjectAlarm(stub);
		expect((await stub.stopFill()).state).toBe("stopped");
		expect(await runDurableObjectAlarm(stub)).toBe(false);
		expect((await stub.page(1, 24)).cached).toBe(true);
	});

	it("a webapp failure fails the job loudly, once: status says why, the log too", async () => {
		const spy = vi.spyOn(console, "error");
		try {
			const stub = await wired("no-such-webapp");
			await stub.startFill();
			await drain(stub);
			const status = await stub.fillStatus();
			expect(status?.state).toBe("failed");
			expect(status?.error).toMatch(/unknown id/);
			expect(spy.mock.calls.map((c) => c[0]).filter((e) => e?.event === "fill")).not.toHaveLength(0);
		} finally {
			spy.mockRestore();
		}
	});

	it("a public cache's fill stores no private rows", async () => {
		const stub = testEnv.PUBLIC_DO.get(testEnv.PUBLIC_DO.newUniqueId());
		await stub.setup({ webappId: "fake-webapp-id" });
		await stub.startFill();
		await drain(stub);
		await runInDurableObject(stub, (_obj, state) => {
			const rows = state.storage.sql.exec<{ n: number }>("select count(*) as n from items where private = 1").toArray();
			expect(rows[0].n).toBe(0);
		});
	});
});

describe("fill routes (goal 1e)", () => {
	const call = (path: string, init?: RequestInit) =>
		worker.fetch(new Request(`https://diigo-cache.test${path}`, init), env as never);
	const post = (body: Record<string, string>) =>
		call("/_fill", { method: "POST", headers: { "content-type": "application/json" }, body: JSON.stringify(body) });

	it("POST /_fill starts the owner's job (cache id in the body) and answers 202 with its status", async () => {
		const id = testEnv.CACHE_DO.newUniqueId();
		await testEnv.CACHE_DO.get(id).setup({ webappId: "fake-webapp-id" });
		const res = await post({ cacheId: toWebkey(id) });
		expect(res.status).toBe(202);
		expect(await res.json()).toMatchObject({ state: "running" });
	});

	it("target=public fills the cache's public companion; action=stop stops", async () => {
		const wiredRes = await call("/_wire", {
			method: "POST",
			headers: { "content-type": "application/json" },
			body: JSON.stringify({ webappId: "fake-webapp-id", mode: "public" }),
		});
		const { cacheId } = (await wiredRes.json()) as { cacheId: string };
		expect((await post({ cacheId, target: "public" })).status).toBe(202);
		const stopped = await post({ cacheId, target: "public", action: "stop" });
		expect(await stopped.json()).toMatchObject({ state: "stopped" });
	});

	it("refuses an unknown cache id (404) and never takes the id from the URL", async () => {
		expect((await post({ cacheId: toWebkey(testEnv.CACHE_DO.newUniqueId()) })).status).toBe(404);
		expect((await call(`/_fill/${toWebkey(testEnv.CACHE_DO.newUniqueId())}`, { method: "POST" })).status).toBe(404);
	});

	it("the created page offers a fetch-everything button, the id in a hidden field", async () => {
		const res = await call("/_wire", {
			method: "POST",
			headers: { "content-type": "application/x-www-form-urlencoded" },
			body: new URLSearchParams({ webappId: "fake-webapp-id", mode: "private" }),
		});
		const html = await res.text();
		expect(html).toMatch(/<form method="post" action="\/_fill">/);
		expect(html).toMatch(/<input type="hidden" name="cacheId"/);
	});

	it("the public page shows a running fill's progress", async () => {
		const wiredRes = await call("/_wire", {
			method: "POST",
			headers: { "content-type": "application/json" },
			body: JSON.stringify({ webappId: "fake-webapp-id", mode: "public" }),
		});
		const { cacheId, username } = (await wiredRes.json()) as { cacheId: string; username: string };
		await post({ cacheId, target: "public" });
		const html = await (await call(`/public/${username}`)).text();
		expect(html).toMatch(/fill(ed|ing)[^<]*\d+ of \d+ pages/);
	});
});