A Shopping List PWA That Learns Your Frequently-Bought Items

Every time you add an item, the app increments a frequency counter for that item name. The top 10 become "frequently bought" suggestions that appear as one-tap chips. Plus: categories, voice input via Web Speech API, multiple lists, and a PWA manifest so you can install it to your phone's home screen.

A shopping list is a simple app. What makes one genuinely useful is learning from your behavior: if you buy milk every week, the app should make adding "milk" a single tap. That's what the frequency-tracking system does.

🔗 Live demo: https://sen.ltd/portfolio/shopping-list/
📦 GitHub: https://github.com/sen-ltd/shopping-list

Features:

9 categories (Produce, Dairy, Meat, Bakery, Pantry, Frozen, Beverage, Household, Other)
Multiple named lists (Weekly, Party, etc.)
Frequently-bought suggestions based on history
Voice input via Web Speech API
PWA manifest (installable)
localStorage persistence
Import / Export JSON
Japanese / English UI
Mobile-first responsive
Zero dependencies, 44 tests

Frequency tracking

Every addItem call increments the history counter:

export function addItem(state, listId, item) {
  const newItem = { ...item, id: generateId(), checked: false, addedAt: Date.now() };
  return {
    ...state,
    lists: state.lists.map(list =>
      list.id === listId ? { ...list, items: [...list.items, newItem] } : list
    ),
    history: {
      ...state.history,
      [item.name.toLowerCase()]: (state.history[item.name.toLowerCase()] || 0) + 1,
    },
  };
}

Normalizing to lowercase means "Milk" and "milk" count as the same item. Over time, the history map grows with one entry per distinct item and a count of how often it's been added.

Getting the top N

export function getFrequentItems(state, limit = 10) {
  return Object.entries(state.history)
    .sort((a, b) => b[1] - a[1])
    .slice(0, limit)
    .map(([name, count]) => ({ name, count }));
}

Simple sort-and-slice. Renders as one-tap chips at the top of the UI: click a suggestion, it gets added to the current list with a default category and quantity.

Voice input with Web Speech API

function startVoiceInput(onResult) {
  const SR = window.SpeechRecognition || window.webkitSpeechRecognition;
  if (!SR) return false;
  const recog = new SR();
  recog.lang = getLang() === 'ja' ? 'ja-JP' : 'en-US';
  recog.continuous = false;
  recog.interimResults = false;
  recog.onresult = (e) => onResult(e.results[0][0].transcript);
  recog.start();
  return true;
}

The API is not supported in Firefox or some Safari versions, so the function returns false and the UI falls back to a text input. When it works, you hold the mic button and say "milk" — the transcript populates the name field.

Category grouping

The rendering groups items by category so you can shop section-by-section in the store:

export function groupByCategory(items) {
  const groups = {};
  for (const item of items) {
    const cat = item.category || 'other';
    if (!groups[cat]) groups[cat] = [];
    groups[cat].push(item);
  }
  return groups;
}

Rendering iterates CATEGORIES in display order so the sections appear consistently even if you add items in a different order.

PWA manifest

{
  "name": "Shopping List",
  "short_name": "Shopping",
  "start_url": "/",
  "display": "standalone",
  "background_color": "#ffffff",
  "theme_color": "#4caf50",
  "icons": [
    { "src": "assets/icon-192.svg", "sizes": "192x192", "type": "image/svg+xml" },
    { "src": "assets/icon-512.svg", "sizes": "512x512", "type": "image/svg+xml" }
  ]
}

Paired with an HTML <link rel="manifest" href="manifest.json">. Chrome on Android offers "Add to Home Screen" and the app launches in standalone mode without browser chrome. For a shopping list this is perfect — it feels native, syncs nothing, stores everything in localStorage.