From 2a3051f2da3e2579c7a2e663d49e5f5b344d8c6d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Andr=C3=A9=20Diamond?=
 <32074058+Andre-Diamond@users.noreply.github.com>
Date: Sat, 27 Apr 2024 08:41:29 +0200
Subject: [PATCH] Refactor meeting summaries retrieval and commit process

---
 .../batchUpdateMeetingSummariesById.js        | 126 ++++++++++--------
 1 file changed, 67 insertions(+), 59 deletions(-)

diff --git a/netlify/functions/batchUpdateMeetingSummariesById.js b/netlify/functions/batchUpdateMeetingSummariesById.js
index bb3fc78..b138587 100644
--- a/netlify/functions/batchUpdateMeetingSummariesById.js
+++ b/netlify/functions/batchUpdateMeetingSummariesById.js
@@ -2,76 +2,84 @@
 import { supabase } from '../../lib/supabaseClient';
 import { Octokit } from "@octokit/rest";
 
-const BATCH_SIZE = 200; // Increase the batch size
-const MAX_CONCURRENT_REQUESTS = 5; // Adjust the number of concurrent requests
+const BATCH_SIZE = 200;
+const MAX_CONCURRENT_REQUESTS = 5;
 
-export const handler = async (event, context) => {
-  try {
-    let allSummaries = {};
-    let lastProcessedTimestamp = null;
-    let hasMoreSummaries = true;
-
-    while (hasMoreSummaries) {
-      const { data: summaries, error } = await supabase
-        .from('meetingsummaries')
-        .select('created_at, meeting_id, summary')
-        .order('created_at', { ascending: true })
-        .limit(BATCH_SIZE * MAX_CONCURRENT_REQUESTS)
-        .gt('created_at', lastProcessedTimestamp || '1970-01-01');
+async function fetchMeetingSummaries(lastProcessedTimestamp) {
+  const { data: summaries, error } = await supabase
+    .from('meetingsummaries')
+    .select('created_at, meeting_id, summary')
+    .order('created_at', { ascending: true })
+    .limit(BATCH_SIZE * MAX_CONCURRENT_REQUESTS)
+    .gt('created_at', lastProcessedTimestamp || '1970-01-01');
 
-      if (error) {
-        console.error('Error retrieving meeting summaries:', error);
-        return {
-          statusCode: 500,
-          body: JSON.stringify({ error: 'Failed to retrieve meeting summaries' }),
-        };
-      }
-
-      if (summaries.length === 0) {
-        hasMoreSummaries = false;
-        break;
-      }
+  if (error) {
+    throw new Error('Failed to retrieve meeting summaries');
+  }
 
-      // Group summaries by meeting_id
-      summaries.forEach(summary => {
-        const { meeting_id, summary: summaryText } = summary;
-        if (!allSummaries[meeting_id]) {
-          allSummaries[meeting_id] = [];
-        }
-        allSummaries[meeting_id].push(summaryText);
-      });
+  return summaries;
+}
 
-      lastProcessedTimestamp = summaries[summaries.length - 1].created_at;
+function groupSummariesByMeetingId(summaries, allSummaries) {
+  summaries.forEach(summary => {
+    const { meeting_id, summary: summaryText } = summary;
+    if (!allSummaries[meeting_id]) {
+      allSummaries[meeting_id] = [];
     }
+    allSummaries[meeting_id].push(summaryText);
+  });
+}
 
-    // Commit all summaries to GitHub in a single file
-    const octokit = new Octokit({
-      auth: process.env.GITHUB_TOKEN,
-    });
-
-    // Get the current SHA of the file
-    let currentSHA = null;
-    try {
-      const { data: currentFile } = await octokit.repos.getContent({
-        owner: "SingularityNET-Archive",
-        repo: "SingularityNET-Archive",
-        path: "Data/Meeting-Summaries/meeting-summaries-by-id.json",
-      });
-      currentSHA = currentFile.sha;
-    } catch (error) {
-      if (error.status !== 404) {
-        throw error;
-      }
-    }
+async function commitSummariesToGitHub(allSummaries) {
+  const octokit = new Octokit({ auth: process.env.GITHUB_TOKEN });
 
-    const { data } = await octokit.repos.createOrUpdateFileContents({
+  let currentSHA = null;
+  try {
+    const { data: currentFile } = await octokit.repos.getContent({
       owner: "SingularityNET-Archive",
       repo: "SingularityNET-Archive",
       path: "Data/Meeting-Summaries/meeting-summaries-by-id.json",
-      message: "Update meeting summaries",
-      content: Buffer.from(JSON.stringify(allSummaries, null, 2)).toString('base64'),
-      sha: currentSHA,
     });
+    currentSHA = currentFile.sha;
+  } catch (error) {
+    if (error.status !== 404) {
+      throw error;
+    }
+  }
+
+  await octokit.repos.createOrUpdateFileContents({
+    owner: "SingularityNET-Archive",
+    repo: "SingularityNET-Archive",
+    path: "Data/Meeting-Summaries/meeting-summaries-by-id.json",
+    message: "Update meeting summaries",
+    content: Buffer.from(JSON.stringify(allSummaries, null, 2)).toString('base64'),
+    sha: currentSHA,
+  });
+}
+
+async function processAndCommitSummaries() {
+  const allSummaries = {};
+  let lastProcessedTimestamp = null;
+  let hasMoreSummaries = true;
+
+  while (hasMoreSummaries) {
+    const summaries = await fetchMeetingSummaries(lastProcessedTimestamp);
+
+    if (summaries.length === 0) {
+      hasMoreSummaries = false;
+      break;
+    }
+
+    groupSummariesByMeetingId(summaries, allSummaries);
+    lastProcessedTimestamp = summaries[summaries.length - 1].created_at;
+  }
+
+  await commitSummariesToGitHub(allSummaries);
+}
+
+export const handler = async (event, context) => {
+  try {
+    await processAndCommitSummaries();
 
     return {
       statusCode: 200,