From 03e2c5dcf27ec69b80e77c527fc8b851f192e432 Mon Sep 17 00:00:00 2001
From: Xavi Lee <awxiaoxian2020@163.com>
Date: Wed, 11 Oct 2023 11:25:43 +0800
Subject: [PATCH] update

---
 src/utilities/fetch-supporters.mjs | 84 +++++++++++++++++++++++++-----
 1 file changed, 72 insertions(+), 12 deletions(-)

diff --git a/src/utilities/fetch-supporters.mjs b/src/utilities/fetch-supporters.mjs
index 3062f24e2531..fd39d0993463 100644
--- a/src/utilities/fetch-supporters.mjs
+++ b/src/utilities/fetch-supporters.mjs
@@ -2,7 +2,6 @@
 import fs from 'fs';
 import path from 'path';
 import { promisify } from 'util';
-import fetch from 'node-fetch';
 import lodash from 'lodash';
 import { fileURLToPath } from 'url';
 
@@ -23,7 +22,22 @@ const absoluteFilename = path.resolve(
   filename
 );
 
-const graphqlEndpoint = 'https://api.opencollective.com/graphql/v2';
+let graphqlEndpoint = 'https://api.opencollective.com/graphql/v2';
+
+if (process.env.OPENCOLLECTIVE_API_KEY) {
+  // rate limit is 100 requests per minute with personal access token
+  // rate limit is 10 requests per minute without personal access token
+  console.log(
+    'Using personal access token to fetch supporters from OpenCollective'
+  );
+  // by default a personal access token of @chenxsan was used as I don't have access to the webpack one
+  // @doc https://graphql-docs-v2.opencollective.com/access#with-a-personal-token
+  graphqlEndpoint = `https://api.opencollective.com/graphql/v2?personalToken=${process.env.OPENCOLLECTIVE_API_KEY}`;
+} else {
+  console.log(
+    'No personal access token found, using public API to fetch supporters from OpenCollective'
+  );
+}
 
 // https://github.com/opencollective/opencollective-api/blob/master/server/graphql/v2/query/TransactionsQuery.ts#L81
 const graphqlPageSize = 1000;
@@ -91,26 +105,72 @@ const getAllNodes = async (graphqlQuery, getNodes) => {
 
   let allNodes = [];
 
+  let limit = 10,
+    remaining = 10,
+    reset;
+  if (process.env.OPENCOLLECTIVE_API_KEY) {
+    limit = 100;
+    remaining = 100;
+  }
   // Handling pagination if necessary
   // eslint-disable-next-line
   while (true) {
+    if (remaining === 0) {
+      console.log(`Rate limit exceeded. Sleeping until ${new Date(reset)}.`);
+      await new Promise((resolve) =>
+        setTimeout(resolve, reset - Date.now() + 100)
+      );
+    }
     const result = await fetch(graphqlEndpoint, {
       method: 'POST',
       body: JSON.stringify(body),
       headers: {
         'Content-Type': 'application/json',
       },
-    }).then((response) => response.json());
-    console.log(result);
-    if (result.errors) throw new Error(result.errors[0].message);
-    const nodes = getNodes(result.data);
-    allNodes = [...allNodes, ...nodes];
-    body.variables.offset += graphqlPageSize;
-    if (nodes.length < graphqlPageSize) {
-      return allNodes;
+    }).then(async (response) => {
+      if (response.headers.get('content-type').includes('json')) {
+        const json = await response.json();
+        console.log('json', json);
+        if (json.error) {
+          // when rate limit exceeded, api won't return headers data like x-ratelimit-limit, etc.
+          remaining = 0;
+          reset = Date.now() + 1000 * 60; // 1 minute
+        } else {
+          limit = response.headers.get('x-ratelimit-limit') * 1;
+          remaining = response.headers.get('x-ratelimit-remaining') * 1;
+          reset = response.headers.get('x-ratelimit-reset') * 1000;
+          console.log(
+            `Rate limit: ${remaining}/${limit} remaining. Reset in ${new Date(
+              reset
+            )}`
+          );
+        }
+        return json;
+      } else {
+        // utilities/fetch-supporters: SyntaxError: Unexpected token < in JSON at position 0
+        console.log('something wrong when fetching supporters');
+        return {
+          error: {
+            message: await response.text(),
+          },
+        };
+      }
+    });
+    // when rate limit exceeded, api will return {error: {message: ''}}
+    // but we could hopefully avoid rate limit by sleeping in the beginning of the loop
+    // however, when there're multiple task running simultaneously, it's still possible to hit the rate limit
+    if (result.error) {
+      console.log('error', result.error);
+      // let the loop continue
     } else {
-      // sleep for a while
-      await new Promise((resolve) => setTimeout(resolve, 6000));
+      const nodes = getNodes(result.data);
+      allNodes = [...allNodes, ...nodes];
+      body.variables.offset += graphqlPageSize;
+      if (nodes.length < graphqlPageSize) {
+        return allNodes;
+      } else {
+        // more nodes to fetch
+      }
     }
   }
 };