## Demo: request-cache ### RequestCache.svelte ```svelte

Each panel calls its own helper, but they all share one memoized analysis — a single parse per request, so one miss and four hits. Open the debug bar (bottom right) and its Cache tab to check the stats.

``` ### analyzeBook.ts ```ts import { requestMemo } from 'mochi-framework'; // The text of Robinson Crusoe, read once at module load and trimmed to Defoe's // prose — the Project Gutenberg header/footer is boilerplate that would skew the // word counts (and hand "unenforceability" the longest-word crown). The file // read is one-time setup; the expensive, repeated work is analyze() below — that // is what gets memoized. const RAW = await Bun.file('./src/demos/request-cache/robinson-crusoe.txt').text(); const bodyStart = RAW.indexOf('***', RAW.indexOf('*** START OF') + 3) + 3; const BOOK = RAW.slice(bodyStart, RAW.indexOf('*** END OF')); // Very common English words carry no signal in a top-words list; drop them so // the narrative's real vocabulary (shore, boat, island, Friday…) rises to the top. const STOPWORDS = new Set( 'a an the and or but if of to in into on upon at by for from with without within as than then so such no not nor only own same other another some any each few many more most all both very much well now here there when where while though yet also about before after over under again ever never how what which who whom whose that this these those i me my we us our you your he him his she her it its they them their been being am is are was were be do does did done have has had having would should could shall will may might must can cannot thus therefore hence unto amongst out up down'.split( ' ', ), ); const THEME_WORDS = ['island', 'sea', 'ship', 'god', 'friday', 'money', 'fear'] as const; export interface Analysis { words: number; unique: number; sentences: number; /** Estimated reading time in minutes at 250 wpm. */ readingMinutes: number; /** All words sorted by descending frequency, stopwords excluded. */ topWords: Array<[string, number]>; themes: Array<[string, number]>; longestSentenceWords: number; longestSentenceExcerpt: string; longestWord: string; /** Count of words appearing exactly once (hapax legomena). */ hapax: number; } /** * The heavy pass: tokenize the whole book, count every word, split sentences, * and derive all five facets in one sweep. ~16–25ms of pure CPU — the unit of * work we want to run once, not once per facet. */ function analyze(text: string): Analysis { const words = text.toLowerCase().match(/[a-z']+/g) ?? []; const freq = new Map(); for (const w of words) { freq.set(w, (freq.get(w) ?? 0) + 1); } const sentences = text.split(/[.!?]+/).filter((s) => s.trim().length > 0); let longestSentenceWords = 0; let longestSentenceExcerpt = ''; for (const s of sentences) { const n = (s.match(/[a-z']+/gi) ?? []).length; if (n > longestSentenceWords) { longestSentenceWords = n; longestSentenceExcerpt = s.replace(/\s+/g, ' ').trim(); } } let longestWord = ''; let hapax = 0; for (const [word, count] of freq) { if (count === 1) { hapax++; } if (word.length > longestWord.length) { longestWord = word; } } const topWords = [...freq.entries()].filter(([w]) => !STOPWORDS.has(w)).sort((a, b) => b[1] - a[1]); const themes = THEME_WORDS.map((w) => [w, freq.get(w) ?? 0] as [string, number]); return { words: words.length, unique: freq.size, sentences: sentences.length, readingMinutes: Math.round(words.length / 250), topWords, themes, longestSentenceWords, longestSentenceExcerpt, longestWord, hapax, }; } /** * The whole-book analysis, memoized for the duration of one request. Zero args * means a single shared entry: however many of the facet helpers below get * called, and from however many components, the book is parsed exactly once per * request — the first call is a miss, every other call is a hit. */ const analyzeCached = requestMemo(() => analyze(BOOK), { namespace: 'demo:crusoe' }); // Five independent facets. Each is its own consumer that asks for the whole // analysis and keeps just its slice — five calls, one parse. export interface Overview { words: number; unique: number; sentences: number; readingMinutes: number; } export function overview(): Overview { const a = analyzeCached(); return { words: a.words, unique: a.unique, sentences: a.sentences, readingMinutes: a.readingMinutes }; } export function topWords(): Array<[string, number]> { return analyzeCached().topWords.slice(0, 12); } export function themes(): Array<[string, number]> { return analyzeCached().themes; } export interface Extremes { longestSentenceWords: number; longestSentenceExcerpt: string; longestWord: string; } export function extremes(): Extremes { const a = analyzeCached(); return { longestSentenceWords: a.longestSentenceWords, longestSentenceExcerpt: a.longestSentenceExcerpt, longestWord: a.longestWord, }; } export function richness(): number { return analyzeCached().hapax; } ``` ### routes.ts ```ts import { Mochi } from 'mochi-framework'; import type { MochiRouteValue } from 'mochi-framework'; export const routes: Record = { '/demos/request-cache': Mochi.page('./src/demos/request-cache/RequestCache.svelte'), }; ``` ### index.ts ```ts import { Mochi, logger } from 'mochi-framework'; import { routes } from './routes'; await Mochi.serve({ port: 3333, development: process.env.MODE === 'development', routes, }); logger.info('Server running at http://localhost:3333'); ```

Overview

Top words

Themes

Extremes

Vocabulary richness