GitHub - moritzrinow/coreblog: Super simple blog engine using S3. No database required.

https://github.com/moritzrinow/coreblog • Nov 24, 2024 18:19

Extracto

Super simple blog engine using S3. No database required. - GitHub - moritzrinow/coreblog: Super simple blog engine using S3. No database required.

Contenido

CoreBlog

CoreBlog is a super simple blog engine, that does not require a database. All you need is an S3 bucket.

Demo: https://coreblog-demo.moritzrinow.com

✅ Multiple themes including dark and white modes
✅ Supports multiple languages including en, de, es, it, fr, uk, ru
✅ SEO friendly
✅ Low resource requirements (100MiB RAM)
✅ No database required
✅ Multiple fonts available
✅ Allows embedding assets in posts
✅ Docker image
✅ Mobile friendly
❌ Sadly no custom UI to edit the blog yet (planned for the future)
- See Manage your blog with Terraform
❌ No real-time blog/post updates (data is pulled in intervals)

Configuration
Install with Docker
S3 bucket layout and file formats
Manage your blog with Terraform

Configuration

coreblog.yaml:

s3:
  endpoint: 'https://fsn1.your-objectstorage.com'
  region: 'fsn1'
  bucketName: 'my-blog' # Required, should be private
  assetBucketName: 'my-blog-assets' # Optional, must be publicly accessible
  accessKey: 'HDY...'
  secretKey: 'r78...'
syncPeriodMinutes: 60 # Every hour blog and post metadata is pulled from S3
postContentCacheTtlMinutes: 60 # Post content is cached in-memory

Install with Docker

Create a coreblog.yaml file with your configurations.

Run the docker container:

docker run -p 5000:5000 -v /my/coreblog.yaml:/coreblog/coreblog.yaml devmojo/coreblog

You can specify a different path for the config file:

docker run devmojo/coreblog --config /path/to/config.yaml

Or specify everything with environment variables:

CB_S3__Endpoint
CB_S3__Region
CB_S3__BucketName
CB_S3__AssetBucketName
CB_S3__AccessKey
CB_S3__SecretKey
CB_SyncPeriodMinutes
CB_PostContentCacheTtlMinutes

S3 bucket layout and file formats

blog.json
posts/
    post-1.json
    post-1.md
    post-2.json
    post-2.md

The S3 bucket stores 3 different kind of files:

1 blog manifest file (blog.json)
- Contains metadata about the blog
N post manifest files (posts/post-id.json)
- Contains metadata about a post
N post content files (posts/post-id.md)
- Contains the post content written in Markdown

Blog manifest (`blog.json`):

{
  "title": "CoreBlog Demo",
  "author": "CoreBlog",
  "description": "Welcome to the CoreBlog demo!",
  "homepage": "https://github.com/moritzrinow/coreblog",
  // Your personal homepage
  "theme": "standard",
  "language": "en",
  // Controls language of UI elements
  "fontFamily": "Roboto Mono",
  "additionalPageMeta": {
    // You can use this to perform things like Google site verifications
  }
}

If the property homepage is specified, a link named 'About Me' will be displayed in the header.

Supported themes:

material
material-dark
standard
standard-dark (used in demo)
default
dark
humanistic
humanistic-dark
software
software-dark

Supported fonts:

Roboto Mono
Roboto
sans-serif

Supported languages:

en (English)
de (German)
es (Spanish)
it (Italian)
fr (French)
uk (Ukrainian)
ru (Russian)

If you want to display a favicon, you have to put a file named favicon.ico in the asset bucket.

Post manifest (`post-id.json`):

{
  "id": "example-1",
  // URL slug, must be same as file prefix
  "title": "How to host your own managed Kubernetes cluster in the cloud",
  "language": "en",
  // Language of the actual content
  "summary": "In this post I will show you how to host your own Kubernetes cluster in the cloud.",
  "published": true,
  "hidden": false,
  // Can be used to have unlisted posts
  "date": "2024-11-21",
  "thumbnail": "welcome.jpg",
  // Image stored in asset bucket
  "tags": [
    "kubernetes",
    "cloud"
  ]
}

Embedding images (assets)

If you want to make use of static assets, such as images, an optional second S3 bucket comes into play.

It must be publicly accessible to serve the assets directly.

To include assets in your blog-post Markdown, use relative links prefixed with assets/:

The following is an image from my assets bucket:
[My Image](assets/my-image.png)

Links starting with assets/ will be rewritten to point to the S3 asset-bucket directly.

Manage your blog with Terraform

Since your whole blog consists of simple files stored in S3, we can make use of the AWS Terraform Provider to streamline the management of our blog.

You can use your IDE of choice to edit your Markdown/JSON files and let Terraform sync that with S3. Same goes for the assets.

Your Terraform module structure:

main.tf
blog.tf
blog-posts/
    post-1.json
    post-1.md
    post-2.json
    post-2.md
blog-assets/
    favicon.ico

Example provider configuration for S3 on Hetzner:

// Provider requirements...

provider "aws" {
  region                      = local.s3_region
  access_key                  = var.s3_access_key
  secret_key                  = var.s3_secret_key
  skip_region_validation      = true
  skip_requesting_account_id  = true
  skip_metadata_api_check     = true
  skip_credentials_validation = true
  endpoints {
    s3 = "https://fsn1.your-objectstorage.com"
  }
}

blog.tf:

// Main bucket
resource "aws_s3_bucket" "blog" {
  bucket = "my-blog"
}

// Asset bucket
resource "aws_s3_bucket" "blog_assets" {
  bucket = "my-blog-assets"
}

// ACL to allow public-read access to asset bucket
resource "aws_s3_bucket_acl" "blog_assets" {
  bucket = aws_s3_bucket.blog.id
  acl    = "public-read"
}

// Our blog manifest (blog.json)
resource "aws_s3_object" "blog_manifest" {
  bucket = aws_s3_bucket.blog.bucket
  key    = "blog.json"
  content = jsonencode({
    title              = "CoreBlog Demo"
    author             = "CoreBlog"
    description        = "Welcome to the CoreBlog demo!"
    homepage           = "https://github.com/moritzrinow/coreblog"
    theme              = "standard-dark"
    language           = "en"
    fontFamily         = "Roboto Mono"
    additionalPageMeta = {}
  })
}

// Post manifest files
resource "aws_s3_object" "blog_posts_manifest" {
  for_each = fileset("${path.module}/blog-posts", "*.json")
  bucket   = aws_s3_bucket.blog.bucket
  key      = "posts/${each.value}"
  source   = "${path.module}/blog-posts/${each.value}"
  etag     = filemd5("${path.module}/blog-posts/${each.value}")
}

// Post content files
resource "aws_s3_object" "blog_posts_content" {
  for_each = fileset("${path.module}/blog-posts", "*.md")
  bucket   = aws_s3_bucket.blog.bucket
  key      = "posts/${each.value}"
  source   = "${path.module}/blog-posts/${each.value}"
  etag     = filemd5("${path.module}/blog-posts/${each.value}")
}

// Blog assets
resource "aws_s3_object" "blog_assets" {
  for_each = fileset("${path.module}/blog-assets", "*")
  bucket   = aws_s3_bucket.blog_assets.bucket
  key      = each.value
  source   = "${path.module}/blog-assets/${each.value}"
  etag     = filemd5("${path.module}/blog-assets/${each.value}")
}

If you modify or add files in the specified directories, Terraform will ensure the specific changes are synced with S3.