<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Satrio Wisnugroho on Medium]]></title>
        <description><![CDATA[Stories by Satrio Wisnugroho on Medium]]></description>
        <link>https://medium.com/@wisnugro?source=rss-9baff034ee23------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*S142gtd7zJiHgvmQ.</url>
            <title>Stories by Satrio Wisnugroho on Medium</title>
            <link>https://medium.com/@wisnugro?source=rss-9baff034ee23------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 09:47:23 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@wisnugro/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Why We Chose Astro Islands Over a Full React SPA]]></title>
            <link>https://medium.com/@wisnugro/why-we-chose-astro-islands-over-a-full-react-spa-d4a0e2bd2f63?source=rss-9baff034ee23------2</link>
            <guid isPermaLink="false">https://medium.com/p/d4a0e2bd2f63</guid>
            <category><![CDATA[web-development]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[react]]></category>
            <category><![CDATA[astrojs]]></category>
            <dc:creator><![CDATA[Satrio Wisnugroho]]></dc:creator>
            <pubDate>Fri, 10 Apr 2026 13:01:02 GMT</pubDate>
            <atom:updated>2026-04-10T13:01:02.025Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QZFioXHbWuxS67Y9BBNo6Q.png" /></figure><p><a href="https://smallhelpertools.com/">Small Helper Tools</a> is a free, browser-based utility site with a growing collection of tools — wheel spinners, dice rollers, finance calculators, text utilities, developer tools, and more. We built it using Astro with React islands, and that architecture decision shaped the entire project.</p><p>In this post, we will explain why we went with islands instead of a traditional React SPA, how we structured the project, and what trade-offs we ran into along the way.</p><h3>The Problem</h3><p>When you have many tools on one site, the JavaScript adds up fast. A traditional React SPA would bundle every tool into one app — the wheel spinner, the loan calculator, the word counter, all of it. Even with code splitting, the router, shared state, and base framework still ship on every page load.</p><p>Most of our pages do not need JavaScript at all. The homepage, category pages, about page, and privacy policy are just static HTML. Only the actual tools — the wheel spinner, dice roller, calculators — need interactivity.</p><p>We did not want users to download a 200KB JavaScript bundle just to read a page with no interactive elements.</p><h3>What Astro Islands Are</h3><p>Astro is a web framework that outputs static HTML by default. No JavaScript ships to the browser unless you explicitly ask for it.</p><p>The “islands” concept is simple. You write your page layout in Astro components, which render to plain HTML. When you need an interactive piece — a React component that handles clicks, state, and animations — you drop it into the page with a client:load directive. Astro calls these interactive pieces “islands” because they are isolated pockets of JavaScript in a sea of static HTML.</p><p>Here is what a typical tool page looks like in our project:</p><pre>import ToolLayout from &#39;../../layouts/ToolLayout.astro&#39;;<br>import WheelSpinner from &#39;../../islands/wheel/WheelSpinner&#39;;<br><br>&lt;ToolLayout<br>  title=&quot;Wheel Spinner&quot;<br>  toolName=&quot;Wheel Spinner&quot;<br>  toolSlug=&quot;wheel-spinner&quot;<br>  category=&quot;random&quot;<br>&gt;<br>  &lt;WheelSpinner client:load /&gt;<br>&lt;/ToolLayout&gt;</pre><p>The layout — navigation, footer, breadcrumbs, SEO content — is all Astro. Zero JavaScript. The WheelSpinner component is a React island that hydrates in the browser and handles all the interactive logic.</p><p>That client:load directive is the key part. Without it, the React component would render on the server as static HTML and never become interactive.</p><h3>How We Structured It</h3><p>We split everything into two clear groups.</p><p><strong>Astro components</strong> handle the layout and structure. These are things like Nav.astro, Footer.astro, Breadcrumb.astro, SEOContent.astro, and ToolCard.astro. They render at build time and produce plain HTML. No client-side JavaScript.</p><p><strong>React islands</strong> handle the interactive tools. Each tool category has its own folder under src/islands/ — wheel/, dice/, finance/, text/, race/, and so on. Shared interactive pieces like the name input field and sound toggle live in islands/shared/.</p><p>The Astro pages in src/pages/ are thin wrappers. Each one imports a layout and a React island, passes some props, and that is it. The page files are typically under 30 lines.</p><p>This structure means adding a new tool does not affect any existing tool. We add an island, create a page, and the rest of the site is untouched.</p><h3>The Real Impact</h3><p>Most page loads on our site ship zero JavaScript. The homepage, category hubs, about page — all static HTML with no JS payload. The page loads fast because there is nothing to download, parse, or execute.</p><p>When a user visits an actual tool — <a href="https://smallhelpertools.com/finance/loan-calculator">say the loan calculator</a> — only that tool’s JavaScript loads. The loan calculator island hydrates, and the rest of the page stays as static HTML.</p><p>This matters more as the site grows. We now have tools across categories like <a href="https://smallhelpertools.com/random">random pickers</a>, <a href="https://smallhelpertools.com/finance">finance calculators</a>, <a href="https://smallhelpertools.com/text">text utilities</a>, <a href="https://smallhelpertools.com/dev-tools">developer tools</a>, and more. In an SPA, every new tool would increase the base bundle. With islands, each tool is independent. The 40th tool does not make the first tool slower.</p><p>The build output is pure static HTML files. No server runtime, no Node process, no serverless functions. We deploy to Cloudflare Pages and it just serves files from a CDN.</p><h3>The Trade-Off: Isolated State</h3><p>The biggest downside of islands is that each one is isolated. You cannot share React state between two islands on the same page. There is no shared context provider wrapping everything, because each island has its own React root.</p><p>For most of our tools, this does not matter. Each tool page has one main island that handles everything.</p><p>Where it did matter was sharing user preferences across tools. For example, the sound toggle — users expect their mute preference to persist across the wheel spinner, dice roller, and race games. We could not use React context for this.</p><p>Our solution was localStorage. The sound preference is stored in localStorage, and each island reads it on mount. It is not as clean as shared React state, but it works and it is simple. We use the same pattern for persisting name lists and calculator inputs.</p><h3>Working With Claude</h3><p>We used Claude throughout the project to write both Astro components and React islands. One thing we noticed: Claude understands the Astro island pattern well and correctly uses client:load directives without being told. It knows which code belongs in an Astro component and which belongs in a React island.</p><p>But we still had to fix things manually. We hit a hydration issue with React 19 where components rendered on the server but silently failed to become interactive in the browser. No error messages, no warnings — just dead buttons. The fix was adding react-dom/client to Vite’s optimizeDeps.include in the Astro config. Claude helped us debug that one too, but it took a few rounds of back-and-forth to narrow it down.</p><p>AI works well for generating consistent patterns. Once we defined the structure for one tool — Astro page wrapper, React island, localStorage persistence — Claude replicated it accurately for every new tool. That saved us a lot of time as the tool count grew.</p><h3>Conclusion</h3><p>Astro islands gave us the best of both worlds. We get fast, static HTML for the pages that do not need interactivity, and full React capabilities for the tools that do. The trade-off — isolated state between islands — was manageable with localStorage.</p><p>If you are building a site with many independent interactive sections surrounded by mostly static content, this architecture is worth considering. The JavaScript savings add up as the page count grows.</p><p>That is it for this one. Hope you found it useful.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d4a0e2bd2f63" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building a Spin Wheel with SVG and CSS Animations in React]]></title>
            <link>https://medium.com/@wisnugro/building-a-spin-wheel-with-svg-and-css-animations-in-react-60b86054b634?source=rss-9baff034ee23------2</link>
            <guid isPermaLink="false">https://medium.com/p/60b86054b634</guid>
            <category><![CDATA[react]]></category>
            <category><![CDATA[svg]]></category>
            <category><![CDATA[web-development]]></category>
            <category><![CDATA[astrojs]]></category>
            <category><![CDATA[javascript]]></category>
            <dc:creator><![CDATA[Satrio Wisnugroho]]></dc:creator>
            <pubDate>Wed, 08 Apr 2026 13:01:02 GMT</pubDate>
            <atom:updated>2026-04-09T09:20:30.758Z</atom:updated>
            <content:encoded><![CDATA[<p>A spin wheel is a circle divided into colored segments, where each segment represents a name or option. The user clicks “Spin”, the wheel rotates with natural deceleration, and a winner is picked at random. We built one for <a href="https://smallhelpertools.com/random/wheel-spinner">Small Helper Tools</a> using React, SVG, and the Web Audio API. The whole thing runs in the browser with no backend.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_xBHVnQ8S_zzEMC0AAMoLQ.png" /><figcaption>Wheel Spinner overview — name input, colorful wheel, and Spin button</figcaption></figure><p>In this post, we will walk through how we drew the wheel, made it spin, added tick sounds, and celebrated the winner with confetti.</p><h3>Drawing the Wheel with SVG</h3><p>The wheel is an SVG element with a 400x400 viewBox and a radius of 180 pixels. Each segment is a pie slice drawn with an SVG path element.</p><p>We use a 12-color palette that cycles through the segments:</p><pre>const COLORS = [<br>  &#39;#6366f1&#39;, &#39;#f59e0b&#39;, &#39;#10b981&#39;, &#39;#ef4444&#39;, &#39;#8b5cf6&#39;, &#39;#ec4899&#39;,<br>  &#39;#14b8a6&#39;, &#39;#f97316&#39;, &#39;#3b82f6&#39;, &#39;#84cc16&#39;, &#39;#e11d48&#39;, &#39;#06b6d4&#39;,<br>];</pre><p>Each segment angle is 360 / names.length. We calculate the start and end points on the circle using basic trigonometry, then draw an SVG arc path:</p><pre>const startAngle = (i * segmentAngle * Math.PI) / 180;<br>const endAngle = ((i + 1) * segmentAngle * Math.PI) / 180;<br>const x1 = center + radius * Math.sin(startAngle);<br>const y1 = center - radius * Math.cos(startAngle);<br>const x2 = center + radius * Math.sin(endAngle);<br>const y2 = center - radius * Math.cos(endAngle);<br>const largeArc = segmentAngle &gt; 180 ? 1 : 0;<br>// Move to center, line to arc start, arc to end, close<br>d={`M${center},${center} L${x1},${y1} A${radius},${radius} 0 ${largeArc},1 ${x2},${y2} Z`}</pre><p>The text label for each segment sits at 65% of the radius from the center, rotated to align with the middle of the slice. We also adjust font size based on how many names are on the wheel — 14px for 6 or fewer, 12px for 7 to 10, and 10px beyond that.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/888/1*r2OaZ0qQnPikp5kKw6As1g.png" /><figcaption>Close-up of the 10-segment wheel showing the color palette and segment labels</figcaption></figure><h3>The Spin Physics</h3><p>When the user clicks “Spin”, we need to pick a winner and rotate the wheel so it lands on the correct segment. Here is how we calculate the rotation.</p><p>First, we pick a random winner index. Then we figure out where in the circle that segment is, add some padding so the pointer does not land right on the edge, and calculate how far the wheel needs to rotate:</p><pre>const winnerIndex = Math.floor(randomFloat() * names.length);<br>const segmentAngle = 360 / names.length;<br>// Land within the segment, not on the edge (10% padding)<br>const padding = segmentAngle * 0.1;<br>const randomOffset = padding + randomFloat() * (segmentAngle - 2 * padding);<br>const segmentPosition = winnerIndex * segmentAngle + randomOffset;<br>const targetAngle = 360 - segmentPosition;</pre><p>To make the spin feel real, we add 5 to 8 extra full rotations before landing at the target. This gives the visual effect of the wheel spinning fast and slowing down:</p><pre>const extraSpins = 5 + Math.floor(randomFloat() * 4);<br>const totalDelta = delta + extraSpins * 360;</pre><p>The actual animation is a CSS transition on the SVG group’s transform property. The spin takes 4 seconds with a custom cubic-bezier curve:</p><pre>transition: transform 4s cubic-bezier(0.2, 0.8, 0.3, 1);</pre><p>The cubic-bezier(0.2, 0.8, 0.3, 1) curve starts with moderate speed, peaks quickly, then decelerates gradually at the end. This feels like a real wheel losing momentum — fast at first, then slower and slower until it stops.</p><h3>Using Cryptographic Randomness</h3><p>For the random number generation, we use crypto.getRandomValues() instead of Math.random(). This matters because people use wheel spinners in classrooms and team meetings where fairness counts.</p><p>Our randomFloat() function is simple:</p><pre>export function randomFloat(): number {<br>  const array = new Uint32Array(1);<br>  crypto.getRandomValues(array);<br>  return array[0] / (0xffffffff + 1);<br>}</pre><p>It generates a cryptographically secure random 32-bit integer and converts it to a float between 0 and 1. We call this for picking the winner, for the offset within the winning segment, and for the number of extra spins. Every random decision in the spin uses the Web Crypto API.</p><h3>Adding Tick Sounds with the Web Audio API</h3><p>A spinning wheel without sound feels lifeless. We wanted a tick sound every time the pointer crosses a segment boundary.</p><p>Every 30 milliseconds during the spin, we check which segment the pointer is currently over. If the segment changed since the last check, we play a tick:</p><pre>tickTimerRef.current = setInterval(checkSegment, 30);</pre><p>Inside the check function, we approximate the current rotation using the same easing curve as the CSS animation:</p><pre>const progress = Math.min(elapsed / 4000, 1);<br>  const eased = 1 - Math.pow(1 - progress, 3);<br>  const currentRot = startRotation + totalDelta * eased;<br>  const currentSegment = Math.floor(((currentRot % 360) + 360) % 360 / segmentAngle);<br>  if (currentSegment !== lastSegment) {<br>  soundManager.playTick();<br>}</pre><p>The tick sound itself is a short sine wave generated with the Web Audio API. No audio file needed:</p><pre>const osc = ctx.createOscillator();<br>const gain = ctx.createGain();<br>osc.type = &#39;sine&#39;;<br>osc.frequency.value = 1200;<br>gain.gain.setValueAtTime(0.2, ctx.currentTime);<br>gain.gain.exponentialRampToValueAtTime(0.001, ctx.currentTime + 0.05);<br>osc.connect(gain);<br>gain.connect(ctx.destination);<br>osc.start();<br>osc.stop(ctx.currentTime + 0.05);</pre><p>This creates a 1200 Hz tone that lasts 50 milliseconds and fades out quickly. The result is a crisp, satisfying click on every segment boundary. The ticks come fast at the start of the spin and slow down naturally as the wheel decelerates.</p><h3>Confetti on Winner Reveal</h3><p>After the 4-second spin animation completes, we trigger canvas-confetti to celebrate the winner:</p><pre>confetti({<br>  particleCount: 100,<br>  spread: 70,<br>  origin: { y: 0.6 },<br>});</pre><p>This fires 100 confetti particles in a 70-degree cone from the center of the page. Combined with a “ding” sound (an 830 Hz tone with a harmonic overtone at 1660 Hz), it makes the winner reveal feel like a small celebration.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*h9N7gIQ-G_6KeDSVEBJViw.png" /><figcaption>Winner reveal — confetti falling and the “Winner: Julia” card below the wheel</figcaption></figure><h3>Persisting the Name List</h3><p>Users should not have to re-enter their names every time they visit. We save the name list to localStorage under the key random.wheel-spinner.names as a JSON array:</p><pre>localStorage.setItem(storageKey, JSON.stringify(parsed));</pre><p>On mount, we load any saved names from storage. If a winner is removed from the list (a common flow for picking multiple winners without repeats), the updated list is saved back immediately. If the list becomes empty, we clean up the storage key entirely.</p><p>We use a lazy useState initializer to load from localStorage on the first render, which avoids the flash of empty state that useEffect can cause.</p><h3>How Claude Helped</h3><p>Claude helped us generate the SVG path calculations for the wheel segments — the trigonometry for positioning text along curved paths is tedious to write by hand. We reviewed every calculation to make sure segments were evenly distributed. The component structure, state management, and sound integration were areas where AI assistance worked well and saved us real time.</p><h3>Try It</h3><p>That covers the main pieces: SVG for drawing, CSS transitions for animation, Web Crypto for fairness, Web Audio for sound, canvas-confetti for celebration, and localStorage for persistence. Each piece is straightforward on its own, and together they make a wheel that feels satisfying to use.</p><p>Try it live at <a href="https://smallhelpertools.com/random/wheel-spinner">smallhelpertools.com/random/wheel-spinner</a>. Hope you found this useful — happy coding!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=60b86054b634" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How We Deployed a Static Site with Many Tools to Cloudflare Pages for Free]]></title>
            <link>https://medium.com/@wisnugro/how-we-deployed-a-static-site-with-many-tools-to-cloudflare-pages-for-free-6575a04b20a2?source=rss-9baff034ee23------2</link>
            <guid isPermaLink="false">https://medium.com/p/6575a04b20a2</guid>
            <category><![CDATA[static-site]]></category>
            <category><![CDATA[web-development]]></category>
            <category><![CDATA[astrojs]]></category>
            <category><![CDATA[cloudflare]]></category>
            <category><![CDATA[javascript]]></category>
            <dc:creator><![CDATA[Satrio Wisnugroho]]></dc:creator>
            <pubDate>Sun, 05 Apr 2026 02:38:21 GMT</pubDate>
            <atom:updated>2026-04-05T02:38:21.546Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xBWnD8P71dxS6a9tK-qhBQ.png" /></figure><p><a href="https://smallhelpertools.com/">Small Helper Tools</a> is a free, browser-based utility site — wheel spinners, dice rollers, finance calculators, text utilities, developer tools, and more. We built it with Astro and React, and the whole thing runs as static HTML with no backend.</p><p>When it came time to deploy, we needed somewhere that handles static sites well and does not cost anything. We chose Cloudflare Pages. The setup was not as smooth as we expected.</p><h3>Why Cloudflare Pages</h3><p>We looked at Netlify, Vercel, and Cloudflare Pages. All three have free tiers that work for static sites.</p><p>We went with Cloudflare for a few reasons. The free tier has unlimited bandwidth, which matters for a site with many pages and assets. Netlify caps bandwidth at 100GB per month on the free plan. Vercel is generous too, but Cloudflare’s global CDN is hard to beat for raw performance.</p><p>For a static site with no server-side rendering, we did not need Vercel’s serverless functions or Netlify’s edge handlers. We just needed a CDN that serves HTML fast. Cloudflare does that well.</p><h3>Build Setup</h3><p>Astro generates static files by default, so the build setup is simple. In the Cloudflare Pages dashboard, we set:</p><ul><li><strong>Build command:</strong> npm run build</li><li><strong>Build output directory:</strong> dist</li></ul><p>Astro outputs everything to dist/ by default, and Cloudflare picks it up from there.</p><p>One thing we had to configure in astro.config.mjs was the build format:</p><pre>export default defineConfig({<br>  site: &#39;https://smallhelpertools.com&#39;,<br>  output: &#39;static&#39;,<br>  trailingSlash: &#39;never&#39;,<br>  build: { format: &#39;file&#39; },<br>});</pre><p>The format: ‘file’ setting generates pages as about.html instead of about/index.html. This works with Cloudflare’s routing, which automatically serves about.html when someone visits /about. Without this, we had issues with trailing slashes and 404s on some paths.</p><h3>Custom Headers</h3><p>Cloudflare Pages supports custom headers through a _headers file in the public directory. We used this for two things: security headers and cache control.</p><p>Here is our full _headers file:</p><pre>/*<br>  X-Frame-Options: DENY<br>  X-Content-Type-Options: nosniff<br>  Referrer-Policy: strict-origin-when-cross-origin<br>  Permissions-Policy: camera=(), microphone=(), geolocation=()<br><br>/_astro/*<br>  Cache-Control: public, max-age=31536000, immutable<br><br>/images/*<br>  Cache-Control: public, max-age=86400<br><br>/sounds/*<br>  Cache-Control: public, max-age=86400<br><br>/favicon.svg<br>  Cache-Control: public, max-age=86400</pre><p>The security headers apply to all routes. X-Frame-Options: DENY prevents our site from being embedded in iframes. X-Content-Type-Options: nosniff stops browsers from guessing content types. Referrer-Policy controls what URL information gets sent to other sites. Permissions-Policy disables camera, microphone, and geolocation since we do not use them.</p><h3>Cache Strategy</h3><p>We split cache control into two tiers.</p><p>Astro generates hashed filenames for JavaScript and CSS bundles in the _astro/ directory. Since the filename changes whenever the content changes, we can cache these aggressively — one year, immutable. Browsers will never re-fetch these files until we deploy a new version with new hashes.</p><p>For images, sounds, and favicons, we use a one-day cache. These files do not have hashed names, so we want browsers to check for updates more frequently without hitting the network on every page load.</p><h3>Redirects</h3><p>We also use a _redirects file for path changes:</p><pre>/index.html / 301<br>/business/invoice-generator /generators/invoice-generator 301<br>/business /generators 301</pre><p>When we moved the invoice generator from /business/ to /generators/, we added 301 redirects so existing links and bookmarks still work. The _redirects file format is simple — source path, destination path, status code.</p><h3>Build Performance</h3><p>Our build takes about 10–15 seconds. Astro is fast because it does not bundle JavaScript for pages that do not need it. Most of our pages are static HTML with zero JavaScript — only the interactive tools (React islands) include JS bundles.</p><p>Cloudflare’s build pipeline picks up the dist/ output and deploys it globally in under a minute. Total time from git push to live site is usually under two minutes.</p><h3>The Claude Part</h3><p>We used Claude to help scaffold the Astro project structure and write the initial deployment configuration. The AI handled the boilerplate well, but we still had to debug the _headers file format manually — Cloudflare’s syntax is specific and Claude did not get it right on the first try.</p><p>The _headers file format looks simple, but spacing matters. The path line must not have leading spaces, and the header lines below it must be indented with exactly two spaces. Claude initially generated it with inconsistent indentation, and Cloudflare silently ignored the malformed rules. No error messages — just headers that did not apply.</p><h3>What We Learned</h3><p>A few takeaways from deploying a multi-page static site to Cloudflare Pages:</p><p>1. build.format: ‘file’<strong> matters.</strong> If your framework generates page/index.html but your host expects page.html (or vice versa), you will get routing issues. Test this early.</p><p>2. <strong>Test your </strong>_headers<strong> file.</strong> Cloudflare does not warn you if the format is wrong. Use the browser DevTools Network tab to verify headers are actually applied.</p><p>3. <strong>Use immutable caching for hashed assets.</strong> If your build tool generates content-hashed filenames, there is no reason not to cache them forever.</p><p>4. <strong>Set up redirects from day one.</strong> Even if you do not have any yet, having the _redirects file in place makes it easy to add them later when you restructure routes.</p><p>5. <strong>Static sites are fast and free.</strong> With no server to manage and no bandwidth limits on Cloudflare’s free tier, the operational cost is zero.</p><p>You can see the deployed result at <a href="https://smallhelpertools.com/">smallhelpertools.com</a>. Every tool runs entirely in the browser — no server, no sign-up, no cost.</p><p>That is it for this one. Hope you found this useful — happy deploying.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6575a04b20a2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Chatbot Simsimi dengan Telegram di Node.js]]></title>
            <link>https://medium.com/skyshidigital/chatbot-simsimi-dengan-telegram-di-node-js-9baca4fe338f?source=rss-9baff034ee23------2</link>
            <guid isPermaLink="false">https://medium.com/p/9baca4fe338f</guid>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[telegram]]></category>
            <category><![CDATA[simsimi]]></category>
            <category><![CDATA[nodejs]]></category>
            <dc:creator><![CDATA[Satrio Wisnugroho]]></dc:creator>
            <pubDate>Sun, 04 Feb 2018 03:01:55 GMT</pubDate>
            <atom:updated>2018-02-06T04:35:37.080Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H7yGEMs1WbJ0Sj8qZekptQ.jpeg" /><figcaption><a href="https://static.politico.com/">https://static.politico.com</a></figcaption></figure><p>Chatbot adalah suatu program yang dirancang untuk melakukan percakapan kepada satu atau lebih orang melalui text maupun audio. Chatbot biasanya digunakan untuk layanan pelanggan dan proses memperoleh informasi.</p><p>Pada artikel kali ini, kita akan membuat chatbot telegram sederhana menggunakan library <a href="https://www.npmjs.com/package/node-telegram-bot-api">node-telegram-bot-api</a>. Kemudian akan kita integrasikan ke <a href="http://www.simsimi.com/otn/chatmode">simsimi</a> agar dapat melakukan percakapan.</p><h3>Persiapan</h3><p>Sebelum membuat script bot, kita perlu mendaftarkan bot kita di telegram dengan cara mencari akun @BotFather</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/561/1*U9xvlfliBE-0n19jo9cIjA.jpeg" /></figure><p>Ketik /newbot untuk membuat bot baru. Lalu masukkan nama bot dan username untuk bot yang akan kita daftarkan.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/551/1*huwDU6tBhlsACLLSN9kh8A.png" /></figure><p>Kita telah mendaftarkan bot Gebetan dengan username gebetan_bot. Setelah berhasil mendaftarkan bot Gebetan, kita akan mendapatkan token untuk mengakses bot API.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/538/1*X2d2WwYeZvL8ycAdFu9FXg.png" /></figure><p>Sampai disini kita sudah berhasil mendaftarkan bot di telegram. Langkah selanjutnya adalah daftar akun baru di simsimi melalui url <a href="http://developer.simsimi.com/signUp">http://developer.simsimi.com/signUp</a> dan buat aplikasi baru.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fEezqQKx4wbo_uLMUmy8AA.png" /></figure><p>Klik View Details, dan kita akan mendapatkan simsimi key untuk mengakses simsimi API.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/797/1*m_rC6wdBN2_sNrgsBo9cvw.png" /></figure><p>Kita telah berhasil mempersiapkan token telegram dan simsimi. Selanjutnya adalah membuat script bot.</p><h3>Membuat Script Bot</h3><p>Siapkan direktori baru dan install library</p><pre>$ yarn add node-telegram-bot-api request</pre><p>Buat file baru bot.js lalu import library</p><pre>const TelegramBot = require(&#39;node-telegram-bot-api&#39;);</pre><p>Masukkan access token bot Gebetan yang sudah kita dapatkan</p><pre>const token = &#39;TELEGRAM_TOKEN&#39;;</pre><p>Inisialisasi bot</p><pre>const bot = new TelegramBot(token, { polling: true });</pre><p>Listen semua message dan mengirimkan kembali ke chat</p><pre>bot.on(&#39;message&#39;, (msg) =&gt; {<br>    const<strong> </strong>chatId = msg.chat.id;<br>    bot.sendMessage(chatId, msg.text.toString());<br>});</pre><p>Berikut adalah kode lengkapnya:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/b8504d061c0bfd90a18bd9345ec577bc/href">https://medium.com/media/b8504d061c0bfd90a18bd9345ec577bc/href</a></iframe><p>Jalankan bot menggunakan perintah node bot.js dan coba mengirimkan chat ke Gebetan</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/552/1*o35GR3x_tM6Znnvr_NtpKg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/1*I_QvQNcFbpdJJH2FOyYnGQ.gif" /></figure><p>Kemudian integrasikan ke simsimi agar bisa menggunakan conversation API milik simsimi. Berikut adalah kode lengkapnya:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/b4dff12816aa5611a66c5b52c372e646/href">https://medium.com/media/b4dff12816aa5611a66c5b52c372e646/href</a></iframe><p>Jalankan lagi bot menggunakan perintah node bot.js dan coba mengirimkan chat ke Gebetan</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/1*D35ESvNL_901H-dzqKF86A.gif" /></figure><p>Kode telegram bot ini bisa di cek ke github dibawah ini.<br>Sekian tutorial kali ini, semoga bermanfaat.</p><p><a href="https://github.com/satriowisnugroho/telegram-chatbot">satriowisnugroho/telegram-chatbot</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9baca4fe338f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/skyshidigital/chatbot-simsimi-dengan-telegram-di-node-js-9baca4fe338f">Chatbot Simsimi dengan Telegram di Node.js</a> was originally published in <a href="https://medium.com/skyshidigital">SkyshiDigital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Membuat Bot Instagram di Node.js Menggunakan Nightmare.js]]></title>
            <link>https://medium.com/skyshidigital/membuat-bot-instagram-di-node-js-menggunakan-nightmare-js-90dd6c6016f?source=rss-9baff034ee23------2</link>
            <guid isPermaLink="false">https://medium.com/p/90dd6c6016f</guid>
            <category><![CDATA[nightmare]]></category>
            <category><![CDATA[nodejs]]></category>
            <category><![CDATA[electron]]></category>
            <category><![CDATA[web-scraping]]></category>
            <category><![CDATA[instagram]]></category>
            <dc:creator><![CDATA[Satrio Wisnugroho]]></dc:creator>
            <pubDate>Tue, 15 Aug 2017 12:22:20 GMT</pubDate>
            <atom:updated>2017-08-18T01:36:08.249Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ItnvTxSS5Gn3HUqm21KUHQ.png" /><figcaption><a href="http://www.nightmarejs.org/">http://www.nightmarejs.org/</a></figcaption></figure><p>Setelah sebelumnya membahas tentang <a href="https://medium.com/skyshidigital/membuat-web-scraper-di-node-js-497c59628042">web scraper </a>, di artikel ini kita masih menggunakan teknik <em>scraping</em> tetapi bukan hanya untuk mengambil informasi melainkan membuat bot Instagram untuk berjualan 🙄.</p><p>Mungkin kita sering melihat komentar yang tidak penting di akun seleb Instagram seperti <a href="https://www.youthmanual.com/post/fun/laugh-out-loud/5-jenis-spam-instagram-yang-paling-nyebelin-sedunia-digital">jual peninggi badan, pelangsing, pemutih</a>, dan masih banyak lagi yang lainnya. Akun toko online tersebut menggunakan penyedia jasa bot untuk melakukan <em>spam</em> terhadap akun selebgram agar dapat meningkatkan pengunjung dan penjualan di toko online mereka.</p><p>Pada kesempatan ini, kita akan membuat bot Instagram sendiri yang dapat melakukan <em>like</em> dan <em>comment</em> berdasarkan <em>query</em> pencarian Instagram menggunakan <a href="http://www.nightmarejs.org/">NightmareJS</a>. Penamaan method yang <em>simple</em> dan lebih mudah dimengerti membuat <em>Library</em> ini sangat mudah untuk diimplementasi. Hanya dengan pengetahuan dasar tentang HTML, kita sudah bisa membuat mimpi buruk untuk web manapun, persis seperti nama dari library ini 😂. Dengan menggunakan <a href="https://electron.atom.io/">electron</a>, library ini memungkinkan kita untuk dapat melihat hasil scraping di GUI.</p><h3>Setup Project</h3><p>Seperti biasa, mari persiapkan direktori dan beberapa <em>package</em> yang dibutuhkan.</p><pre>yarn add nightmare</pre><h3>Let’s Scrape It !</h3><p>Deklarasi:</p><pre>const Nightmare = require(&#39;nightmare&#39;);<br>const nightmare = Nightmare({<br>    show: true,<br>    webPreferences: {<br>        partition: &#39;nopersist&#39;,<br>    },<br>});<br><br>const username = &#39;wisnugro&#39;;<br>const password = &#39;12345678&#39;;<br>const search = &#39;#jogja&#39;;<br>const comment = &#39;keren&#39;;<br>const posts = 10;</pre><p>Dimulai dengan login ke web Instagram:</p><pre>nightmare<br>    .goto(&#39;http://instagram.com&#39;)<br>    .wait(&#39;._b93kq&#39;)<br>    .click(&#39;._b93kq&#39;) // log in option<br>    .wait(&#39;input[name=username]&#39;)<br>    .insert(&#39;input[name=username]&#39;, username)<br>    .insert(&#39;input[name=password]&#39;, password)<br>    .click(&#39;button&#39;);</pre><p>Lakukan pencarian berdasarkan <em>query</em>:</p><pre>nightmare<br>    .wait(&#39;input[placeholder=Search]&#39;)<br>    .insert(&#39;input[placeholder=Search]&#39;, search)<br>    .wait(&#39;._gimca&#39;)<br>    .click(&#39;._gimca&#39;) // top suggest</pre><p><em>Like</em> dan <em>comment</em> postingan sebanyak yang diinginkan:</p><pre>nightmare<br>    .wait(&#39;._mck9w&#39;)<br>    .click(&#39;._mck9w a&#39;) // click post<br>    .wait(&#39;._eszkz&#39;);<br><br>for (let i = 0; i &lt; posts; i++) {<br>    nightmare.wait(1000)<br>        .click(&#39;._eszkz&#39;) // like<br>        .insert(&#39;._bilrf&#39;, comment) // comment<br>        .type(&#39;._bilrf&#39;, &#39;\u000d&#39;)<br>        .click(&#39;._3a693&#39;); // next<br>}<br><br>nightmare<br>    .wait(1000)<br>    .end()<br>    .then(function () {<br>        console.log(&#39;success&#39;);<br>    })<br>    .catch(function (err) {<br>        console.log(err);<br>    });</pre><p>Berikut merupakan keseluruhan kode:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/aad1ea409d36e4781c2391dde8f6c756/href">https://medium.com/media/aad1ea409d36e4781c2391dde8f6c756/href</a></iframe><h3>Hasil</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*ocz4xgqzroRvM2pJazHfnw.gif" /><figcaption>electron</figcaption></figure><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FuJjlZTLq8jg%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DuJjlZTLq8jg&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FuJjlZTLq8jg%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="640" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/b6cfcb3f17a99465437b435793313cb3/href">https://medium.com/media/b6cfcb3f17a99465437b435793313cb3/href</a></iframe><p>Penggunaan <em>selector</em> <em>class</em> di web Instagram sangat tidak disarankan, karena <em>class</em> akan di <em>generate</em> terus menerus oleh Instagram. Untuk itu cek kode di bawah untuk penggunaan <em>selector</em> <em>tag</em> untuk mendapatkan nama <em>class</em>.</p><p><a href="https://github.com/satriowisnugroho/instagram-auto-likes-comments/blob/master/routes/index.js">satriowisnugroho/instagram-auto-likes-comments</a></p><p>Kunjungi juga link dibawah untuk mencoba bot Instagram.</p><p><a href="https://instagram.satriowisnugroho.me/">Home * Satrio Wisnugroho</a></p><p>Sekian untuk tutorial kali ini, semoga bermanfaat.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=90dd6c6016f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/skyshidigital/membuat-bot-instagram-di-node-js-menggunakan-nightmare-js-90dd6c6016f">Membuat Bot Instagram di Node.js Menggunakan Nightmare.js</a> was originally published in <a href="https://medium.com/skyshidigital">SkyshiDigital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Membuat Web Scraper di Node.js]]></title>
            <link>https://medium.com/skyshidigital/membuat-web-scraper-di-node-js-497c59628042?source=rss-9baff034ee23------2</link>
            <guid isPermaLink="false">https://medium.com/p/497c59628042</guid>
            <category><![CDATA[node-js-tutorial]]></category>
            <category><![CDATA[nodejs]]></category>
            <category><![CDATA[web-scraping]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[scraping]]></category>
            <dc:creator><![CDATA[Satrio Wisnugroho]]></dc:creator>
            <pubDate>Wed, 26 Apr 2017 17:43:14 GMT</pubDate>
            <atom:updated>2021-02-15T03:36:56.418Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DB5Ht-spRa7zuJkoBW46dQ.jpeg" /><figcaption>Node.js Technology (Source: <a href="http://mindinmotion.co/nodejs.html">http://mindinmotion.co/nodejs.html</a>)</figcaption></figure><p>Web scraping adalah proses mengekstrak informasi dari suatu website. Web scraping dilakukan karena data yang dibutuhkan tidak tersedia di API, atau bahkan mereka tidak menyediakan API sama sekali. Selain untuk mengekstrak informasi, web scraping juga digunakan untuk mengotomatisasi proses yang dilakukan secara manual, atau yang bisa disebut dengan bot.</p><p>Pada kesempatan kali ini, kita akan membuat web scraper sederhana untuk menampilkan jadwal sholat selama sebulan.</p><p>Sebelumnya saya sudah membuat project serupa yang bisa dilihat di bawah ini.</p><p><a href="https://github.com/satriowisnugroho/azan">satriowisnugroho/azan</a></p><h3><strong>Setup Project</strong></h3><p>Pertama, kita mulai dengan membuat direktori dan menginstall beberapa package yang dibutuhkan.</p><pre>mkdir azan &amp;&amp; cd azan &amp;&amp; npm install --save request cheerio colors</pre><p>Lalu buat file baru yang kita namakan index.js.</p><p>Selanjutnya tambahkan modul yang sudah kita install sebelumnya:</p><pre>const request = require(&#39;request&#39;);<br>const cheerio = require(&#39;cheerio&#39;);</pre><p>Sampai disini kita sudah siap untuk memulai proses scraping.</p><p><a href="https://www.codementor.io/johnnyb/how-to-write-a-web-scraper-in-nodejs-du108266t">Ada 3 langkah</a> yang dibutuhkan untuk melakukan scraping:</p><ol><li>Load HTML</li><li>Parsing HTML</li><li>Extract Data</li></ol><h3>Load</h3><p>Untuk mendapatkan HTML dari suatu halaman, kita perlu melakukan request ke URL yang akan kita scrape.</p><pre>let url = &#39;https://jadwalsholat.org/adzan/monthly.php?id=307&#39;;</pre><pre>request(url, function (err, res, body) {<br>  if (err &amp;&amp; res.statusCode !== 200) throw err;<br>  console.log(body);<br>});</pre><h3>Parse</h3><p>Setalah mendapatkan HTML dari hasil request tadi, lalu parsing menggunakan <a href="https://github.com/cheeriojs/cheerio">cheerio</a>.</p><pre>let $ = cheerio.load(body);</pre><pre>console.log($(&#39;body&#39;).html());</pre><h3>Extract</h3><p>Sebelum mengekstrak data yang kita butuhkan, kita perlu mempelajari struktur HTML dari halaman tesebut.</p><p>Contohnya kita ingin mendapatkan list jadwal sholat. Kita dapat melakukan inspect element pada browser seperti dibawah ini:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Vpk21VJwXiJIGdp6UIU-dA.jpeg" /></figure><p>Setelah mengetahui tag, class, atau id mana yang berisi data yang kita butuhkan, selanjutnya kita bisa mengekstrak data tersebut.</p><p>Berikut merupakan isi dari file index.js:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/7833f25a75523b45949189185db21619/href">https://medium.com/media/7833f25a75523b45949189185db21619/href</a></iframe><p>Jalankan aplikasi dengan perintah node index.js.</p><p>Berikut merupakan output dari proses scraping yang sudah kita lakukan di atas:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/576/1*4UkR3Gz227DTXoFCxRpsLA.jpeg" /></figure><h3>Kesimpulan</h3><p>Tujuan dari tutorial ini adalah untuk menjelaskan bagaimana menggunakan teknik web scraping untuk mendapatkan data dari website dan menggunakannya sesuai kebutuhan. Akan tetapi kita juga harus memperhatikan <a href="https://towardsdatascience.com/ethics-in-web-scraping-b96b18136f01">kode etik</a> dalam melakukan web scraping.</p><p>Sekian tutorial kali ini, semoga bermanfaat.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=497c59628042" width="1" height="1" alt=""><hr><p><a href="https://medium.com/skyshidigital/membuat-web-scraper-di-node-js-497c59628042">Membuat Web Scraper di Node.js</a> was originally published in <a href="https://medium.com/skyshidigital">SkyshiDigital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Caching menggunakan Redis + Node.js]]></title>
            <link>https://medium.com/skyshidigital/caching-menggunakan-redis-node-js-8b83eb2ed507?source=rss-9baff034ee23------2</link>
            <guid isPermaLink="false">https://medium.com/p/8b83eb2ed507</guid>
            <category><![CDATA[nodejs]]></category>
            <category><![CDATA[redis]]></category>
            <category><![CDATA[cache]]></category>
            <category><![CDATA[javascript]]></category>
            <dc:creator><![CDATA[Satrio Wisnugroho]]></dc:creator>
            <pubDate>Sun, 02 Apr 2017 15:15:56 GMT</pubDate>
            <atom:updated>2017-04-07T14:42:02.043Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DGu_UMsfQyQqFDewHz7t_Q.jpeg" /></figure><p>Untuk membangun suatu API, kita dapat menggunakan cache agar dapat meningkatkan kecepatan, mengurangi penggunaan bandwith dan server load. <br>Pada kesempatan kali ini kita akan membuat API untuk mendapatkan total public repository Github seseorang.</p><h3>Setup Project</h3><p>Kita akan memulai dengan membuat direktori kosong untuk project kita.</p><pre>mkdir node_redis &amp;&amp; cd node_redis</pre><p>Setelah itu install express, redis, dan superagent menggunakan NPM.</p><pre>npm init &amp;&amp; npm install express redis superagent --save</pre><p>Sampai disini kita telah berhasil melakukan instalasi pada project API.</p><h3>Membangun API Server</h3><p>Saatnya kita membuat API server dan membuat satu endpoint untuk mengakses API Github.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/1217e7dcade1e0b6d29ee3a73c485855/href">https://medium.com/media/1217e7dcade1e0b6d29ee3a73c485855/href</a></iframe><p>Jalankan aplikasi dan buat satu request.</p><pre><a href="http://localhost:3000/users?username=satriowisnugroho">http://localhost:3000/users?username=satriowisnugroho</a></pre><p>Pada contoh di bawah ini saya menggunakan Postman. Anda juga bisa menggunakan browser kesayangan Anda untuk melakukan request.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/823/1*wsr4lDLvVj-UNl-7XlAKnA.jpeg" /></figure><p>Saat kita mengakses endpoint users, terlihat kecepatan request timenya yaitu 1363 ms. Jika Anda menggunakan browser, Anda dapat melihat request timenya pada tab network di developer tools.</p><p>Oke, sampai disini kita sudah berhasil membangun API untuk melihat total public repository Github. Selanjutnya kita akan menggunakan caching untuk mempersingkat request timenya.</p><h3>Caching Data</h3><p>Sekarang saatnya kita menggunakan redis untuk menyimpan data secara sementara.</p><pre>client.setex(&#39;key&#39;, 3600, &#39;value&#39;);</pre><p>Setelah menyimpan data, selanjutnya adalah mengambil data cache.</p><pre>client.get(&#39;key&#39;, function (err, data) {});</pre><p>Berikut merupakan isi dari file setelah menggunakan redis.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/65ec76a3175fa4b7fd2172627c6bbe7f/href">https://medium.com/media/65ec76a3175fa4b7fd2172627c6bbe7f/href</a></iframe><p>Restart aplikasi kemudian reload.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/754/1*tHpssuw6YaIgeBAZ7csKow.jpeg" /></figure><p>Terlihat kecepatannya menjadi 15 ms, jauh lebih cepat daripada sebelum menggunakan cache.</p><h3>Kesimpulan</h3><p>Dengan menggunakan cache kita dapat meningkatkan kecepatan akses data. Dengan pendekatan sederhana yang sudah kita coba di atas, kita dapat melihat hasil yang memuaskan. Kita juga dapat mengaplikasikannya pada proses akses data ke database, ataupun ke sumber lain yang menggunakan pemrosesan lebih lambat.</p><p>Oke sekian tutorial singkat dari saya, semoga bermanfaat.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8b83eb2ed507" width="1" height="1" alt=""><hr><p><a href="https://medium.com/skyshidigital/caching-menggunakan-redis-node-js-8b83eb2ed507">Caching menggunakan Redis + Node.js</a> was originally published in <a href="https://medium.com/skyshidigital">SkyshiDigital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>