Sidebar (replaces the modal drawer): - Inline, toggleable, no backdrop/focus theft — can stay open while working - Open state persists across pages (provider in root layout) and reloads (localStorage; starts closed server-side to avoid hydration mismatch, stored value applied post-mount; default open on first run) - Hamburger in page headers only when closed; close button in sidebar header Floating composer blob (per reference screenshot): - Floats over the transcript (absolute overlay), persists at bottom while scrolling, content fades behind it (gradient mask); pb-44 scroll clearance - Concentric corner radii: blob rounded-2xl(16px) − p-1.5(6px) = send rounded-[10px] = 25% of the 40px button; CornerDownLeft icon per reference - Stop moved from header into the blob (busy → square icon) - prompt-input.tsx + 7 orphan ui files pruned (attachments/command palette unused); hand-rolled Composer is ~100 lines and render-tested Pending-question morph (one text field on screen, never two): - Quiz/ask-select pending → blob becomes the note field (send hidden, hint line above); card's internal note input removed; Submit consumes it - Ask free-text pending → blob becomes the answer field; its send submits via the card's registered submit api (ref bridge); Enter works - Chat owns the single text state; card clears it on submit - canSend for answer mode mirrors the card predicate directly — reading the ref during render lagged one commit (found in browser testing) Widths: one shared column (max-w-3xl + px-4) for transcript, cards, and composer — all three measured pixel-identical (left 416, width 736). Assistant Message max-w-full/MessageContent w-full (was max-w-[95%]/w-fit — the 'cards are a bit shorter' bug). User messages render plain text again — no math pipeline on learner-authored text (currency hazard accepted for the model only, guarded by the prompt). Tests: 58 pass (5 new composer pins; card tests updated to controlled props). Verified live: note flow, answer flow, sidebar persistence, float, fade, zero console errors — plus the live model spontaneously honoring the $ currency escape contract in a continuation.
168 lines
7.2 KiB
TypeScript
168 lines
7.2 KiB
TypeScript
/**
|
|
* Render smoke tests — the lesson from the v2 bugs: protocol tests don't see
|
|
* rendering. These render the interaction cards to static HTML and assert on
|
|
* the output. No DOM needed (react-dom/server).
|
|
*
|
|
* tsx note: tsx ignores tsconfig's `jsx` field and uses the CLASSIC JSX
|
|
* transform (React.createElement), so `React` must be a resolvable global.
|
|
* The shim below provides it; Vite/Next use the automatic runtime and don't
|
|
* care.
|
|
*/
|
|
|
|
import React from "react";
|
|
(globalThis as Record<string, unknown>).React = React;
|
|
|
|
import assert from "node:assert/strict";
|
|
import { describe, it } from "node:test";
|
|
import { renderToStaticMarkup } from "react-dom/server";
|
|
import { MessageResponse } from "../components/ai-elements/message.tsx";
|
|
import { QuizCard } from "../components/quiz-card.tsx";
|
|
import { AnsweredCard } from "../components/answered-card.tsx";
|
|
import { buildAnswer, coerceCorrectAnswer, normalizeOptions, resolveCorrect, type QuestionData } from "../src/core/question-core.ts";
|
|
|
|
const quizInput = {
|
|
question: "Which planet is closest to the sun?",
|
|
options: [{ label: "Venus" }, { label: "Mercury" }, { label: "Mars" }],
|
|
correctAnswer: "Mercury",
|
|
explanation: "Mercury is the innermost planet.",
|
|
};
|
|
|
|
function quizQD(): QuestionData {
|
|
const options = normalizeOptions(quizInput.options);
|
|
return {
|
|
id: "q1",
|
|
kind: "quiz",
|
|
question: quizInput.question,
|
|
mode: "single-select",
|
|
options,
|
|
correctIndices: resolveCorrect(coerceCorrectAnswer(quizInput.correctAnswer), options).indices,
|
|
explanation: quizInput.explanation,
|
|
createdAt: new Date().toISOString(),
|
|
};
|
|
}
|
|
|
|
describe("render smoke tests", () => {
|
|
it("QuizCard renders question, options, and the automatic 'I don't know'", () => {
|
|
const html = renderToStaticMarkup(
|
|
<QuizCard toolName="quiz" toolCallId="q1" input={quizInput} onSubmit={() => {}} pendingText="" onPendingTextChange={() => {}} />,
|
|
);
|
|
assert.match(html, /Which planet is closest to the sun\?/);
|
|
assert.match(html, /Mercury/);
|
|
assert.match(html, /I don't know|I don't know/);
|
|
// The v1 desync bug: options are buttons with ARIA roles, not native inputs.
|
|
assert.ok(!html.includes('type="radio"'), "no native radio inputs");
|
|
assert.ok(html.includes('role="radio"'), "button-based radio role present");
|
|
});
|
|
|
|
it("QuizCard never leaks the explanation or the correct answer before submit", () => {
|
|
const html = renderToStaticMarkup(
|
|
<QuizCard toolName="quiz" toolCallId="q1" input={quizInput} onSubmit={() => {}} pendingText="" onPendingTextChange={() => {}} />,
|
|
);
|
|
assert.ok(!html.includes("innermost"), "explanation must not render pre-answer");
|
|
});
|
|
|
|
it("ask card renders Other, no note field (note lives in the composer), no grading", () => {
|
|
const html = renderToStaticMarkup(
|
|
<QuizCard
|
|
toolName="ask_user_question"
|
|
toolCallId="a1"
|
|
input={{ question: "How confident are you?", options: [{ label: "Very" }, { label: "Somewhat" }] }}
|
|
onSubmit={() => {}}
|
|
pendingText=""
|
|
onPendingTextChange={() => {}}
|
|
/>,
|
|
);
|
|
assert.match(html, /Other/);
|
|
assert.ok(!html.includes("I don't know"), "ask has no 'I don't know'");
|
|
assert.ok(!html.includes("Add a note"), "note field moved to the composer blob");
|
|
});
|
|
|
|
it("AnsweredCard grades quiz answers (✓/✗ + explanation)", () => {
|
|
const answer = buildAnswer(quizQD(), { values: ["Venus"] });
|
|
const html = renderToStaticMarkup(<AnsweredCard answer={answer} />);
|
|
assert.match(html, /incorrect/);
|
|
assert.match(html, /innermost/, "explanation renders post-answer");
|
|
assert.ok(html.includes("✓") && html.includes("✗"));
|
|
});
|
|
|
|
it("AnsweredCard does NOT grade ask answers", () => {
|
|
const qd: QuestionData = {
|
|
id: "a1",
|
|
kind: "ask",
|
|
question: "How confident are you?",
|
|
mode: "single-select",
|
|
options: normalizeOptions([{ label: "Very" }, { label: "Somewhat" }]),
|
|
createdAt: new Date().toISOString(),
|
|
};
|
|
const answer = buildAnswer(qd, { values: ["Somewhat"] });
|
|
const html = renderToStaticMarkup(<AnsweredCard answer={answer} />);
|
|
assert.ok(!html.includes("incorrect") && !html.includes("✗"), "ask answers never get quiz grading UI");
|
|
assert.match(html, /Somewhat/);
|
|
});
|
|
|
|
it("AnsweredCard renders dismissal as skipped", () => {
|
|
const answer = { ...buildAnswer(quizQD(), { dontKnow: true }), dismissed: true };
|
|
const html = renderToStaticMarkup(<AnsweredCard answer={answer} />);
|
|
assert.match(html, /skipped/);
|
|
});
|
|
});
|
|
|
|
describe("math rendering (streamdown + KaTeX)", () => {
|
|
it("renders INLINE math ($...$) — the LLM-native convention", () => {
|
|
const html = renderToStaticMarkup(<MessageResponse>{"The map $f(x) = x^2$ is smooth."}</MessageResponse>);
|
|
assert.ok(html.includes("katex"), `inline math should render as KaTeX, got: ${html.slice(0, 300)}`);
|
|
});
|
|
|
|
it("renders display math ($$ fenced on own lines — the prompt contract)", () => {
|
|
const html = renderToStaticMarkup(<MessageResponse>{"$$\n\\int_0^1 x^2\\,dx = \\tfrac{1}{3}\n$$"}</MessageResponse>);
|
|
assert.ok(html.includes("katex-display"), "fenced display math should render as display KaTeX");
|
|
});
|
|
|
|
it("one-line $$...$$ degrades to inline (micromark behavior — prompt forbids it)", () => {
|
|
// micromark treats a one-line $$...$$ inside a paragraph as 2-dollar
|
|
// inline mathText, not flow math. The system prompt instructs the fenced
|
|
// form, so the model never emits this; pin the behavior so a renderer
|
|
// swap can't silently change it.
|
|
const html = renderToStaticMarkup(<MessageResponse>{"$$\\int_0^1 x^2\\,dx$$"}</MessageResponse>);
|
|
assert.ok(html.includes("katex") && !html.includes("katex-display"), "one-liner should render inline");
|
|
});
|
|
|
|
it("does NOT eat escaped currency (\\$5 stays text — the prompt contract)", () => {
|
|
// micromark-extension-math@3.1 lacks pandoc's closing-digit rule, so bare
|
|
// "$5 and $10" would parse as math. The defense is at the source: the
|
|
// system prompt instructs the teacher to escape currency (\$5 / 5 USD).
|
|
const html = renderToStaticMarkup(<MessageResponse>{"It costs \\$5 and \\$10 more."}</MessageResponse>);
|
|
assert.ok(!html.includes("katex"), "escaped currency must not parse as math");
|
|
assert.match(html, /\$5/);
|
|
});
|
|
|
|
it("quiz cards render math in question, options, and explanation", () => {
|
|
const mathQuiz = {
|
|
question: "What is $\\frac{dy}{dx}$ of $y = x^2$?",
|
|
options: [{ label: "$2x$" }, { label: "$x^2$" }, { label: "$x$" }],
|
|
correctAnswer: "$2x$",
|
|
explanation: "The power rule gives $\\frac{dy}{dx} = 2x$.",
|
|
};
|
|
const pendingHtml = renderToStaticMarkup(
|
|
<QuizCard toolName="quiz" toolCallId="m1" input={mathQuiz} onSubmit={() => {}} pendingText="" onPendingTextChange={() => {}} />,
|
|
);
|
|
assert.ok(pendingHtml.includes("katex"), "pending card: question/options math renders");
|
|
|
|
const options = normalizeOptions(mathQuiz.options);
|
|
const qd: QuestionData = {
|
|
id: "m1",
|
|
kind: "quiz",
|
|
question: mathQuiz.question,
|
|
mode: "single-select",
|
|
options,
|
|
correctIndices: resolveCorrect(coerceCorrectAnswer(mathQuiz.correctAnswer), options).indices,
|
|
explanation: mathQuiz.explanation,
|
|
createdAt: new Date().toISOString(),
|
|
};
|
|
const answeredHtml = renderToStaticMarkup(
|
|
<AnsweredCard answer={buildAnswer(qd, { values: ["$2x$"] })} />,
|
|
);
|
|
assert.ok(answeredHtml.includes("katex"), "answered card: option/explanation math renders");
|
|
});
|
|
});
|