From d63f8916670d7c8889aa547225007f829fd83e66 Mon Sep 17 00:00:00 2001
From: Alicia Sykes <alicia@omg.lol>
Date: Sat, 12 Aug 2023 21:57:24 +0100
Subject: [PATCH] Adds API endpoint to fetch data from Wayback Machine

---
 api/archives.js | 83 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 83 insertions(+)
 create mode 100644 api/archives.js

diff --git a/api/archives.js b/api/archives.js
new file mode 100644
index 0000000..1e9ab27
--- /dev/null
+++ b/api/archives.js
@@ -0,0 +1,83 @@
+const axios = require('axios');
+const middleware = require('./_common/middleware');
+
+const convertTimestampToDate = (timestamp) => {
+  const [year, month, day, hour, minute, second] = [
+    timestamp.slice(0, 4),
+    timestamp.slice(4, 6) - 1,
+    timestamp.slice(6, 8),
+    timestamp.slice(8, 10),
+    timestamp.slice(10, 12),
+    timestamp.slice(12, 14),
+  ].map(num => parseInt(num, 10));
+
+  return new Date(year, month, day, hour, minute, second);
+}
+
+const countPageChanges = (results) => {
+  let prevDigest = null;
+  return results.reduce((acc, curr) => {
+    if (curr[2] !== prevDigest) {
+      prevDigest = curr[2];
+      return acc + 1;
+    }
+    return acc;
+  }, -1);
+}
+
+const getAveragePageSize = (scans) => {
+    const totalSize = scans.map(scan => parseInt(scan[3], 10)).reduce((sum, size) => sum + size, 0);
+    return Math.round(totalSize / scans.length);
+};
+
+const getScanFrequency = (firstScan, lastScan, totalScans, changeCount) => {
+  const formatToTwoDecimal = num => parseFloat(num.toFixed(2));
+
+  const dayFactor = (lastScan - firstScan) / (1000 * 60 * 60 * 24);  
+  const daysBetweenScans = formatToTwoDecimal(dayFactor / totalScans);
+  const daysBetweenChanges = formatToTwoDecimal(dayFactor / changeCount);
+  const scansPerDay = formatToTwoDecimal((totalScans - 1) / daysBetweenScans);
+  const changesPerDay = formatToTwoDecimal(changeCount / daysBetweenScans);
+  return {
+    daysBetweenScans,
+    daysBetweenChanges,
+    scansPerDay,
+    changesPerDay,
+  };
+};
+
+const getWaybackData = async (url) => {
+  const cdxUrl = `https://web.archive.org/cdx/search/cdx?url=${url}&output=json&fl=timestamp,statuscode,digest,length,offset`;
+
+  try {
+    const { data } = await axios.get(cdxUrl);
+    
+    // Check there's data
+    if (!data || !Array.isArray(data) || data.length <= 1) {
+      return { skipped: 'Site has never before been archived via the Wayback Machine' };
+    }
+
+    // Remove the header row
+    data.shift();
+
+    // Process and return the results
+    const firstScan = convertTimestampToDate(data[0][0]);
+    const lastScan = convertTimestampToDate(data[data.length - 1][0]);
+    const totalScans = data.length;
+    const changeCount = countPageChanges(data);
+    return {
+      firstScan,
+      lastScan,
+      totalScans,
+      changeCount,
+      averagePageSize: getAveragePageSize(data),
+      scanFrequency: getScanFrequency(firstScan, lastScan, totalScans, changeCount),
+      scans: data,
+      scanUrl: url,
+    };
+  } catch (err) {
+    return { error: `Error fetching Wayback data: ${err.message}` };
+  }
+};
+
+exports.handler = middleware(getWaybackData);