From f7ae9bbde1084e875cb1e7fe23d7d9b089b349e1 Mon Sep 17 00:00:00 2001
From: Kshitiz Shakya <kshitiz.shakya22@gmail.com>
Date: Wed, 14 May 2025 17:14:46 +0545
Subject: [PATCH] Using larger github runner + Move models.json to s3 bucket
 [skip-test]

---
 .github/workflows/create_search_index.yml |  7 ++++++-
 docs/Gemfile                              |  3 +++
 docs/Gemfile.lock                         | 22 ++++++++++++++++++++++
 docs/_plugins/search_index.rb             | 23 +++++++++++++++++++++--
 docs/latest.html                          |  3 ---
 5 files changed, 52 insertions(+), 6 deletions(-)
 delete mode 100644 docs/latest.html

diff --git a/.github/workflows/create_search_index.yml b/.github/workflows/create_search_index.yml
index fece33dc41e8f0..5d4f6db3d2f871 100644
--- a/.github/workflows/create_search_index.yml
+++ b/.github/workflows/create_search_index.yml
@@ -11,7 +11,8 @@ concurrency:
 
 jobs:
   jekyll:
-    runs-on: ubuntu-latest
+    runs-on: RAM32GB
+    timeout-minutes: 600
     environment: jekyll
     steps:
       - uses: actions/checkout@v2
@@ -49,6 +50,8 @@ jobs:
           ELASTICSEARCH_INDEX_NAME: ${{ secrets.ELASTICSEARCH_INDEX_NAME }}
           SEARCH_ORIGIN: ${{ secrets.SEARCH_ORIGIN }}
           ORIGIN: ${{ secrets.ORIGIN }}
+          AWS_ACCESS_KEY_ID: ${{ secrets.MODELS_PUBLIC_KEY }}
+          AWS_SECRET_ACCESS_KEY: ${{ secrets.MODELS_SECRET_KEY }}
         working-directory: docs
         run: |
           bundle exec jekyll build --incremental
@@ -62,6 +65,8 @@ jobs:
           ELASTICSEARCH_INDEX_NAME: ${{ secrets.ELASTICSEARCH_INDEX_NAME }}
           SEARCH_ORIGIN: ${{ secrets.SEARCH_ORIGIN }}
           ORIGIN: ${{ secrets.ORIGIN }}
+          AWS_ACCESS_KEY_ID: ${{ secrets.MODELS_PUBLIC_KEY }}
+          AWS_SECRET_ACCESS_KEY: ${{ secrets.MODELS_SECRET_KEY }}
         working-directory: docs
         run: |
           rm -f .jekyll-metadata
diff --git a/docs/Gemfile b/docs/Gemfile
index 41e93b1c165a29..8a4e98f169603c 100644
--- a/docs/Gemfile
+++ b/docs/Gemfile
@@ -11,6 +11,9 @@ gem "webrick"
 
 gem "jekyll", "~> 3.9"
 
+gem "aws-sdk-s3", "~>1"
+
+
 group "jekyll-plugins" do
     gem "jekyll-incremental", "0.1.0", path: "_plugins/jekyll-incremental"
 end
diff --git a/docs/Gemfile.lock b/docs/Gemfile.lock
index 880e734f5d046b..02bbee56ff83ba 100644
--- a/docs/Gemfile.lock
+++ b/docs/Gemfile.lock
@@ -15,6 +15,25 @@ GEM
       zeitwerk (~> 2.2, >= 2.2.2)
     addressable (2.8.1)
       public_suffix (>= 2.0.2, < 6.0)
+    aws-eventstream (1.4.0)
+    aws-partitions (1.1126.0)
+    aws-sdk-core (3.226.2)
+      aws-eventstream (~> 1, >= 1.3.0)
+      aws-partitions (~> 1, >= 1.992.0)
+      aws-sigv4 (~> 1.9)
+      base64
+      jmespath (~> 1, >= 1.6.1)
+      logger
+    aws-sdk-kms (1.106.0)
+      aws-sdk-core (~> 3, >= 3.225.0)
+      aws-sigv4 (~> 1.5)
+    aws-sdk-s3 (1.192.0)
+      aws-sdk-core (~> 3, >= 3.225.0)
+      aws-sdk-kms (~> 1)
+      aws-sigv4 (~> 1.5)
+    aws-sigv4 (1.12.1)
+      aws-eventstream (~> 1, >= 1.0.2)
+    base64 (0.3.0)
     coffee-script (2.4.1)
       coffee-script-source
       execjs
@@ -233,6 +252,7 @@ GEM
       gemoji (~> 3.0)
       html-pipeline (~> 2.2)
       jekyll (>= 3.0, < 5.0)
+    jmespath (1.6.2)
     kramdown (2.3.2)
       rexml
     kramdown-parser-gfm (1.1.0)
@@ -241,6 +261,7 @@ GEM
     listen (3.8.0)
       rb-fsevent (~> 0.10, >= 0.10.3)
       rb-inotify (~> 0.9, >= 0.9.10)
+    logger (1.7.0)
     mercenary (0.3.6)
     mini_portile2 (2.8.1)
     minima (2.5.1)
@@ -301,6 +322,7 @@ PLATFORMS
   x86_64-linux
 
 DEPENDENCIES
+  aws-sdk-s3 (~> 1)
   elasticsearch (~> 7.10)
   github-pages (= 227)
   jekyll (~> 3.9)
diff --git a/docs/_plugins/search_index.rb b/docs/_plugins/search_index.rb
index 4fc1b38bd71bd8..f8f5139cdabd11 100644
--- a/docs/_plugins/search_index.rb
+++ b/docs/_plugins/search_index.rb
@@ -5,7 +5,9 @@
 require 'date'
 require 'elasticsearch'
 require 'nokogiri'
+require 'aws-sdk-s3'
 
+BUCKET_NAME="pypi.johnsnowlabs.com"
 SEARCH_URL = (ENV["SEARCH_ORIGIN"] || 'https://search.modelshub.johnsnowlabs.com') + '/'
 ELASTICSEARCH_INDEX_NAME = ENV["ELASTICSEARCH_INDEX_NAME"] || 'models'
 
@@ -17,6 +19,18 @@
 
 $remote_editions = Set.new
 
+def upload_file_to_s3_bucket(file_path)
+  s3 = Aws::S3::Client.new(region: 'eu-west-1')
+  object_key = "public/models.json"
+  begin
+    s3.put_object(bucket: BUCKET_NAME, key: object_key, body: File.open(file_path, 'rb'), acl: 'public-read')
+    puts "File uploaded successfully to #{BUCKET_NAME}/#{object_key}"
+
+  rescue Aws::S3::Errors::ServiceError => e
+    puts "Failed to upload file: #{e.message}"
+  end
+end
+
 class Version < Array
   def initialize name
     m = /(\d+\.\d+)\z/.match(name)
@@ -252,7 +266,7 @@ def initialize(client)
 
   def index(id, data)
     @buffer << { update: { _id: id, data: {doc: data, doc_as_upsert: true}} }
-    self.execute if @buffer.length >= 100
+    self.execute if @buffer.length >= 500
   end
 
   def execute
@@ -578,9 +592,14 @@ def is_latest?(group, model)
     models_references_json = backup_references_data.merge(models_references_json)
   end
 
-  filename = File.join(site.config['destination'], 'models.json')
+  filename = File.join(site.config['destination'], 'backup-modelss3.json')
+
   File.write(filename, models_json.values.to_json)
   File.write(backup_filename, models_json.to_json)
+  # models.json moved to pypi s3 bucket
+  upload_file_to_s3_bucket(filename)
+
+  File.delete(filename)
 
   benchmarking_filename = File.join(site.config['destination'], 'benchmarking.json')
   File.write(benchmarking_filename, models_benchmarking_json.to_json)
diff --git a/docs/latest.html b/docs/latest.html
deleted file mode 100644
index 89dc810eadb76b..00000000000000
--- a/docs/latest.html
+++ /dev/null
@@ -1,3 +0,0 @@
----
-layout: archive
----