FluentCMS

Welcome to FluentCMS! 🚀

FluentCMS makes content management seamless with its powerful GraphQL API and intuitive drag-and-drop page design features.

If you'd like to contribute, please check out our CONTRIBUTING guide.
Enjoying FluentCMS? Don’t forget to give us a ⭐ and help us grow!

What is Fluent CMS?

Fluent CMS is an open-source Content Management System designed to simplify and accelerate web development workflows. While it's particularly suited for CMS projects, it is also highly beneficial for general web applications, reducing the need for repetitive REST/GraphQL API development.

• Effortless CRUD Operations: Fluent CMS includes built-in RESTful APIs for Create, Read, Update, and Delete (CRUD) operations, complemented by a React-based admin panel for efficient data management.

• Powerful GraphQL Queries: Access multiple related entities in a single query, enhancing client-side performance, security, and flexibility.

• Drag-and-Drop Page Designer: Build dynamic pages effortlessly using the integrated page designer powered by Grapes.js and Handlebars. Easily bind data sources for an interactive and streamlined design experience.

Online Course System Demo

Source Code

Example Project on GitHub

Live Demo

• Public Site: fluent-cms-admin.azurewebsites.net
• Admin Panel: fluent-cms-admin.azurewebsites.net/admin
  • Email: [e-mail]
  • Password: Admin1!

Additional Resources

• GraphQL Playground: fluent-cms-admin.azurewebsites.net/graph
• Documentation: fluent-cms-admin.azurewebsites.net/doc/index.html

Online Course System Backend

This section provides detailed guidance on developing a foundational online course system, encompassing key entities: teacher, course, lesson,skill, and material.

Database Schema

1. Teachers Table

The Teachers table maintains information about instructors, including their personal and professional details.

Field	Header	Data Type
id	ID	Int
firstname	First Name	String
lastname	Last Name	String
email	Email	String
phone_number	Phone Number	String
image	Image	String
bio	Bio	Text

2. Courses Table

The Courses table captures the details of educational offerings, including their scope, duration, and prerequisites.

Field	Header	Data Type
id	ID	Int
name	Course Name	String
status	Status	String
level	Level	String
summary	Summary	String
image	Image	String
desc	Description	Text
duration	Duration	String
start_date	Start Date	Datetime

3. Lessons Table

The Lessons table contains detailed information about the lessons within a course, including their title, content, and associated teacher.

Field	Header	Data Type
id	ID	Int
name	Lesson Name	String
description	Description	Text
teacher	Teacher	Int (Foreign Key)
course	Course	Int (Foreign Key)
created_at	Created At	Datetime
updated_at	Updated At	Datetime

4. Skills Table

The Skills table records competencies attributed to teachers.

Field	Header	Data Type
id	ID	Int
name	Skill Name	String
years	Years of Experience	Int
created_by	Created By	String
created_at	Created At	Datetime
updated_at	Updated At	Datetime

5. Materials Table

The Materials table inventories resources linked to courses.

Field	Header	Data Type
id	ID	Int
name	Name	String
type	Type	String
image	Image	String
link	Link	String
file	File	String

Relationships

• Courses to Teachers: Man-to-One(Each teacher can teach multiple courses; each course is assigned to one teacher. A teacher can exist independently of a course).
• Teachers to Skills: Many-to-Many (Multiple teachers can share skills, and one teacher may have multiple skills).
• Courses to Materials: Many-to-Many (A course may include multiple materials, and the same material can be used in different courses).
• Courses to Lessons: One-to-Many (Each course can have multiple lessons, but each lesson belongs to one course. A lesson cannot exist without a course, as it has no meaning on its own).

Schema Creation via Fluent CMS Schema Builder

Accessing Schema Builder

After launching the web application, locate the Schema Builder menu on the homepage to start defining your schema.

Adding Entities

Example Configuration

• Navigate to the Entities section of the Schema Builder.
• Create entities such as "Teacher" and "Course."
• For the Course entity, add attributes such as name, status, level, and description.

Defining Relationships

Example Configuration

1. Course and Teacher (Many-to-One Relationship)

To establish a many-to-one relationship between the Course and Teacher entities, you can include a Lookup attribute in the Course entity. This allows selecting a single Teacher record when adding or updating a Course.

Attribute	Value
Field	teacher
DataType	Lookup
DisplayType	Lookup
Options	Teacher

Description: When a course is created or modified, a teacher record can be looked up and linked to the course.

2 ** Course and Lesson(One-to-Many Relationship)**

To establish a one-to-many relationship between the Course and Lesson entities, use a Collection attribute in the Course entity. This enables associating multiple lessons with a single course.

Attribute	Value
Field	lessons
DataType	Collection
DisplayType	EditTable
Options	Lesson

Description: When managing a course , you can manage lessons of this course.

3. Course and Materials (Many-to-Many Relationship)

To establish a many-to-many relationship between the Course and Material entities, use a Junction attribute in the Course entity. This enables associating multiple materials with a single course.

Attribute	Value
Field	materials
DataType	Junction
DisplayType	Picklist
Options	Material

Description: When managing a course, you can select multiple material records from the Material table to associate with the course.

Admin Panel

The last chapter introduced how to model entities, this chapter introduction how to use Admin-Panel to manage data of those entities.

Display Types

The Admin Panel supports various UI controls to display attributes:

• "text": Single-line text input.

• "textarea": Multi-line text input.

• "editor": Rich text input.

• "number": Single-line text input for numeric values only.

• "datetime": Datetime picker for date and time inputs.

• "date": Date picker for date-only inputs.

• "image": Upload a single image, storing the image URL.

• "gallery": Upload multiple images, storing their URLs.

• "file": Upload a file, storing the file URL.

• "dropdown": Select an item from a predefined list.

• "multiselect": Select multiple items from a predefined list.

• "lookup": Select an item from another entity with a many-to-one relationship (requires Lookup data type).

• "picklist": Select multiple items from another entity with a many-to-many relationship (requires Junction data type).

• "edittable": Manage items of a one-to-many sub-entity (requires Collection data type).

DataType to DisplayType Mapping Table

Below is a mapping of valid DataType and DisplayType combinations:

DataType	DisplayType	Description
Int	Number	Input for integers.
Datetime	Datetime	Datetime picker for date and time inputs.
Datetime	Date	Date picker for date-only inputs.
String	Number	Input for numeric values.
String	Datetime	Datetime picker for date and time inputs.
String	Date	Date picker for date-only inputs.
String	Text	Single-line text input.
String	Textarea	Multi-line text input.
String	Image	Single image upload.
String	Gallery	Multiple image uploads.
String	File	File upload.
String	Dropdown	Select an item from a predefined list.
String	Multiselect	Select multiple items from a predefined list.
Text	Multiselect	Select multiple items from a predefined list.
Text	Gallery	Multiple image uploads.
Text	Textarea	Multi-line text input.
Text	Editor	Rich text editor.
Lookup	Lookup	Select an item from another entity.
Junction	Picklist	Select multiple items from another entity.
Collection	EditTable	Manage items of a sub-entity.

List Page

Example Course List Page

The List Page displays entities in a tabular format, supporting sorting, searching, and pagination for efficient browsing or locating of specific records.

Sorting

Sort records by clicking the ↑ or ↓ icon in the table header.

• Order by Created At Example
• Order by Name Example

Filtering

Apply filters by clicking the Funnel icon in the table header.

• Filter by Created At (2024-09-07)
• Filter by Course Name (Starts with A or C)

Detail Page

Detail page provides an interface to manage single record.

Example of display types date,image, gallery, muliselect, dropdown,

Lesson Detail Page.

Example of lookup,picklist,edittable

Course Detail Page

GraphQL Query

FluentCMS simplifies frontend development by offering robust GraphQL support.

Getting Started

Accessing the GraphQL IDE

To get started, launch the web application and navigate to /graph. You can also try our online demo.

Singular vs. List Response

For each entity in FluentCMS, two GraphQL fields are automatically generated:

• <entityName>: Returns a record.
• <entityNameList>: Returns a list of records.

**Single Course **

{
  course {
    id
    name
  }
}

Try it here

**List of Courses **

{
  courseList {
    id
    name
  }
}

Try it here

Field Selection

You can query specific fields for both the current entity and its related entities. Example Query:

{
  courseList{
    id
    name
    teacher{
      id
      firstname
      lastname
      skills{
        id
        name
      }
    }
    materials{
      id,
      name
    }
  }
}

Try it here

Filtering with Value Match in FluentCMS

FluentCMS provides flexible filtering capabilities using the idSet field (or any other field), enabling precise data queries by matching either a single value or a list of values.

Filter by a Single Value Example:

{
  courseList(idSet: 5) {
    id
    name
  }
}

Try it here

Filter by Multiple Values Example:

{
  courseList(idSet: [5, 7]) {
    id
    name
  }
}

Try it here

Advanced Filtering with Operator Match in FluentCMS

FluentCMS supports advanced filtering options with Operator Match, allowing users to combine various conditions for precise queries.

matchAll Example:

Filters where all specified conditions must be true.
In this example: id > 5 and id < 15.

{
  courseList(id: {matchType: matchAll, gt: 5, lt: 15}) {
    id
    name
  }
}

Try it here

matchAny Example:

Filters where at least one of the conditions must be true.
In this example: name starts with "A" or name starts with "I".

{
  courseList(name: [{matchType: matchAny}, {startsWith: "A"}, {startsWith: "I"}]) {
    id
    name
  }
}

Try it here

Filter Expressions in FluentCMS

Filter Expressions allow precise filtering by specifying a field, including nested fields using JSON path syntax. This enables filtering on subfields for complex data structures.

Example: Filter by Teacher's Last Name This query returns courses taught by a teacher whose last name is "Circuit."

{
  courseList(filterExpr: {field: "teacher.lastname", clause: {equals: "Circuit"}}) {
    id
    name
    teacher {
      id
      lastname
    }
  }
}

Try it here

Sorting

Sorting by a single field

{
  courseList(sort:nameDesc){
    id,
    name
  }
}

Try it here

Sorting by multiple fields

{
  courseList(sort:[level,id]){
    id,
    level
    name
  }
}

Try it here

Sort Expressions in FluentCMS

Sort Expressions allow sorting by nested fields using JSON path syntax.

Example: Sort by Teacher's Last Name

{
  courseList(sortExpr:{field:"teacher.lastname", order:Desc}) {
    id
    name
    teacher {
      id
      lastname
    }
  }
}

Try it here

Pagination

Pagination on root field

{
  courseList(offset:2, limit:3){
    id,
    name
  }
}

Try it here
Pagination on sub field

{
  courseList{
    id,
    name
    materials(limit:2){
      id,
      name
    }
  }
}

Try it here

Variable

Variables are used to make queries more dynamic, reusable, and secure.

Variable in Value filter

query ($id: Int!) {
  teacher(idSet: [$id]) {
    id
    firstname
    lastname
  }
}

Try it here

Variable in Operator Match filter

query ($id: Int!) {
  teacherList(id:{equals:$id}){
    id
    firstname
    lastname
  }
}

Try it here

Variable in Filter Expression

query ($years: String) {
  teacherList(filterExpr:{field:"skills.years",clause:{gt:$years}}){
    id
    firstname
    lastname
    skills{
      id
      name
      years
    }
  }
}

Try it here

Variable in Sort

query ($sort_field:TeacherSortEnum) {
  teacherList(sort:[$sort_field]) {
    id
    firstname
    lastname
  }
}

Try it here

Variable in Sort Expression

query ($sort_order:  SortOrderEnum) {
  courseList(sortExpr:{field:"teacher.id", order:$sort_order}){
    id,
    name,
    teacher{
      id,
      firstname
    }
  }
}

Try it here

Variable in Pagination

query ($offset:Int) {
  teacherList(limit:2, offset:$offset) {
    id
    firstname
    lastname
  }
}

Try it here

Required vs Optional

If you want a variable to be mandatory, you can add a ! to the end of the type

query ($id: Int!) {
  teacherList(id:{equals:$id}){
    id
    firstname
    lastname
  }
}

Try it here

Explore the power of FluentCMS GraphQL and streamline your development workflow!

Saved Query

Realtime queries may expose excessive technical details, potentially leading to security vulnerabilities.

Saved Queries address this issue by abstracting the GraphQL query details. They allow clients to provide only variables, enhancing security while retaining full functionality.

Transitioning from Real-Time Queries to Saved Queries

Using OperationName as the Saved Query Identifier

In FluentCMS, the Operation Name in a GraphQL query serves as a unique identifier for saved queries. For instance, executing the following query automatically saves it as TeacherQuery:

query TeacherQuery($id: Int) {
  teacherList(idSet: [$id]) {
    id
    firstname
    lastname
    skills {
      id
      name
    }
  }
}

Try it here

Saved Query Endpoints

FluentCMS generates two API endpoints for each saved query:

• List Records:
  https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery

• Single Record:
  https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery/single/

Using REST API Query Strings as Variables

The Saved Query API allows passing variables via query strings:

• Single Value:
  https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery/?id=3

• Array of Values:
  https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery?id=3&id=4
  This passes [3, 4] to the idSet argument.

Additional Features of Saved Query

Beyond performance and security improvements, Saved Query introduces enhanced functionalities to simplify development workflows.

Pagination by offset

Built-in variables offset and limit enable efficient pagination. For example:
https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery?limit=2&offset=2

offset Pagination for Subfields

To display a limited number of subfield items (e.g., the first two skills of a teacher), use the JSON path variable, such as skills.limit:
https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery?skills.limit=2

Pagination by cursor

For large datasets, offset pagination can strain the database. For example, querying with offset=1000&limit=10 forces the database to retrieve 1010 records and discard the first 1000.

To address this, Saved Query supports cursor-based pagination, which reduces database overhead.
Example response for https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery?limit=3:

[
  {
    "hasPreviousPage": false,
    "cursor": "eyJpZCI6M30"
  },
  {
  },
  {
    "hasNextPage": true,
    "cursor": "eyJpZCI6NX0"
  }
]

• If hasNextPage of the last record is true, use the cursor to retrieve the next page:
  https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery?limit=3&last=eyJpZCI6NX0

• Similarly, if hasPreviousPage of the first record is true, use the cursor to retrieve the previous page:
  https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery?limit=3&first=eyJpZCI6Nn0

Cursor-Based Pagination for Subfields

Subfields also support cursor-based pagination. For instance, querying https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery?skills.limit=2 returns a response like this:

[
  {
    "id": 3,
    "firstname": "Jane",
    "lastname": "Debuggins",
    "hasPreviousPage": false,
    "skills": [
      {
        "hasPreviousPage": false,
        "cursor": "eyJpZCI6MSwic291cmNlSWQiOjN9"
      },
      {
        "hasNextPage": true,
        "cursor": "eyJpZCI6Miwic291cmNlSWQiOjN9"
      }
    ],
    "cursor": "eyJpZCI6M30"
  }
]

To fetch the next two skills, use the cursor:
https://fluent-cms-admin.azurewebsites.net/api/queries/TeacherQuery/part/skills?limit=2&last=eyJpZCI6Miwic291cmNlSWQiOjN9

Drag and Drop Page Designer

The page designer utilizes the open-source GrapesJS and Handlebars, enabling seamless binding of GrapesJS Components with FluentCMS Queries for dynamic content rendering.

Page Types: Landing Page, Detail Page, and Home Page

Landing Page

A landing page is typically the first page a visitor sees.

• URL format: /page/<pagename>
• Structure: Comprised of multiple sections, each section retrieves data via a query.

Example:
Landing Page
This page fetches data from:

• https://fluent-cms-admin.azurewebsites.net/api/queries/course/?status=featured
• https://fluent-cms-admin.azurewebsites.net/api/queries/course/?level=Advanced

Detail Page

A detail page provides specific information about an item.

• URL format: /page/<pagename>/<router parameter>
• Data Retrieval: FluentCMS fetches data by passing the router parameter to a query.

Example:
Course Detail Page
This page fetches data from:
https://fluent-cms-admin.azurewebsites.net/api/queries/course/one?course_id=22

Home Page

The homepage is a special type of landing page named home.

• URL format: /pages/home
• Special Behavior: If no other route matches the path /, FluentCMS renders /pages/home by default.

Example:
The URL / will be resolved to /pages/home unless explicitly overridden.

Introduction to GrapesJS Panels

Understanding the panels in GrapesJS is crucial for leveraging FluentCMS's customization capabilities in the Page Designer UI. This section explains the purpose of each panel and highlights how FluentCMS enhances specific areas to streamline content management and page design.

• Style Manager:

  • Used to customize CSS properties of elements selected on the canvas.
  • FluentCMS Integration: This panel is left unchanged by FluentCMS, as it already provides powerful styling options.

• Traits Panel:

  • Allows modification of attributes for selected elements.
  • FluentCMS Integration: Custom traits are added to this panel, enabling users to bind data to components dynamically.

• Layers Panel:

  • Displays a hierarchical view of elements on the page, resembling a DOM tree.
  • FluentCMS Integration: While FluentCMS does not alter this panel, it’s helpful for locating and managing FluentCMS blocks within complex page designs.

• Blocks Panel:

  • Contains pre-made components that can be dragged and dropped onto the page.
  • FluentCMS Integration: FluentCMS enhances this panel by adding custom-designed blocks tailored for its CMS functionality.

By familiarizing users with these panels and their integration points, this chapter ensures a smoother workflow and better utilization of FluentCMS's advanced page-building tools.

Data Binding: Singleton or List

FluentCMS leverages Handlebars expressions for dynamic data binding in pages and components.

Singleton

Singleton fields are enclosed within {{ }} to dynamically bind individual values.

• Example Page Settings: Page Schema Settings
• Example Query: Retrieve Course Data
• Example Rendered Page: Rendered Course Page

List

Handlebars supports iterating over arrays using the {{#each}} block for repeating data structures.

{{#each course}}
    <li>{{title}}</li>
{{/each}}

In FluentCMS, you won’t explicitly see the {{#each}} statement in the Page Designer. If a block's data source is set to data-list, FluentCMS automatically generates the loop.

• Example Page Settings: Page Schema Settings
• Example Rendered Page: Rendered List Page
• Example Queries:
  • Featured Courses
  • Advanced Level Courses

Steps to Bind a Data Source

To bind a Data List to a component, follow these steps:

• Drag a block from the Data List category in the Page Designer.
• Open the Layers Panel and select the Data List component.
• In the Traits Panel, configure the following fields:

Field	Description
Query	The query to retrieve data.
Qs	Query string parameters to pass (e.g., ?status=featured, ?level=Advanced).
Offset	Number of records to skip.
Limit	Number of records to retrieve.
Pagination	Options for displaying content:
	- Button: Divides content into multiple pages with navigation buttons (e.g., "Next," "Previous," or numbered buttons).
	- Infinite Scroll: Automatically loads more content as users scroll. Ideal for a single component at the bottom of the page.
	- None: Displays all available content at once without requiring additional user actions.

Online Course System Frontend

Having established our understanding of Fluent CMS essentials like Entity, Query, and Page, we're ready to build a frontend for an online course website.

Key Pages

• Home Page (home): The main entry point, featuring sections like Featured Courses and Advanced Courses. Each course links to its respective Course Details page.
• Course Details (course/{course_id}): Offers detailed information about a specific course and includes links to the Teacher Details page.
• Teacher Details (teacher/{teacher_id}): Highlights the instructor’s profile and includes a section displaying their latest courses, which link back to the Course Details page.

             Home Page
                 |
                 |
       +-------------------+
       |                   |
       v                   v
 Latest Courses       Course Details 
       |                   |        
       |                   |       
       v                   v            
Course Details <-------> Teacher Details 

Designing the Home Page

• Drag and Drop Components: Use the Fluent CMS page designer to drag a Content-B component.
• Set Data Source: Assign the component's data source to the course query.
• Link Course Items: Configure the link for each course to /pages/course/{{id}}. The Handlebars expression {{id}} is dynamically replaced with the actual course ID during rendering.

Creating the Course Details Page

• Page Setup: Name the page course/{course_id} to capture the course_id parameter from the URL (e.g., /pages/course/20).
• Query Configuration: The variable {course_id:20} is passed to the course query, generating a WHERE id IN (20) clause to fetch the relevant course data.
• Linking to Teacher Details: Configure the link for each teacher item on this page to /pages/teacher/{{teacher.id}}. Handlebars dynamically replaces {{teacher.id}} with the teacher’s ID. For example, if a teacher object has an ID of 3, the link renders as /pages/teacher/3.

Creating the Teacher Details Page

• Page Setup: Define the page as teacher/{teacher_id} to capture the teacher_id parameter from the URL.
• Set Data Source: Assign the teacher query as the page’s data source.

Adding a Teacher’s Courses Section

• Drag a ECommerce A component onto the page.
• Set its data source to the course query, filtered by the teacher’s ID (WHERE teacher IN (3)).

When rendering the page, the PageService automatically passes the teacher_id (e.g., {teacher_id: 3}) to the query.

Optimizing Caching

Fluent CMS employs advanced caching strategies to boost performance.

For detailed information on ASP.NET Core caching, visit the official documentation: ASP.NET Core Caching Overview.

Cache Schema

Fluent CMS automatically invalidates schema caches whenever schema changes are made. The schema cache consists of two types:

• Entity Schema Cache
  Caches all entity definitions required to dynamically generate GraphQL types.

• Query Schema Cache
  Caches query definitions, including dependencies on multiple related entities, to compose efficient SQL queries.

By default, schema caching is implemented using IMemoryCache. However, you can override this by providing a HybridCache. Below is a comparison of the two options:

IMemoryCache

• Advantages:
  • Simple to debug and deploy.
  • Ideal for single-node web applications.
• Disadvantages:
  • Not suitable for distributed environments. Cache invalidation on one node (e.g., Node A) does not propagate to other nodes (e.g., Node B).

HybridCache

• Key Features:
  • Scalability: Combines the speed of local memory caching with the consistency of distributed caching.
  • Stampede Resolution: Effectively handles cache stampede scenarios, as verified by its developers.
• Limitations:
  The current implementation lacks "Backend-Assisted Local Cache Invalidation," meaning invalidation on one node does not instantly propagate to others.
• Fluent CMS Strategy:
  Fluent CMS mitigates this limitation by setting the local cache expiration to 20 seconds (one-third of the distributed cache expiration, which is set to 60 seconds). This ensures cache consistency across nodes within 20 seconds, significantly improving upon the typical 60-second delay in memory caching.

To implement a HybridCache, use the following code:

builder.AddRedisDistributedCache(connectionName: CmsConstants.Redis);
builder.Services.AddHybridCache();

Cache Data

Fluent CMS does not automatically invalidate data caches. Instead, it leverages ASP.NET Core's output caching for a straightforward implementation. Data caching consists of two types:

• Query Data Cache
  Caches the results of queries for faster access.

• Page Cache
  Caches the output of rendered pages for quick delivery.

By default, output caching is disabled in Fluent CMS. To enable it, configure and inject the output cache as shown below:

builder.Services.AddOutputCache(cacheOption =>
{
    cacheOption.AddBasePolicy(policyBuilder => policyBuilder.Expire(TimeSpan.FromMinutes(1)));
    cacheOption.AddPolicy(CmsOptions.DefaultPageCachePolicyName,
        b => b.Expire(TimeSpan.FromMinutes(2)));
    cacheOption.AddPolicy(CmsOptions.DefaultQueryCachePolicyName,
        b => b.Expire(TimeSpan.FromSeconds(1)));
});

// After builder.Build();
app.UseOutputCache();

Aspire Integration

Fluent CMS leverages Aspire to simplify deployment.

Architecture Overview

A scalable deployment of Fluent CMS involves multiple web application nodes, a Redis server for distributed caching, and a database server, all behind a load balancer.

                 +------------------+
                 |  Load Balancer   |
                 +------------------+
                          |
        +-----------------+-----------------+
        |                                   |
+------------------+              +------------------+
|    Web App 1     |              |    Web App 2     |
|   +-----------+  |              |   +-----------+  |
|   | Local Cache| |              |   | Local Cache| |
+------------------+              +------------------+
        |                                   |
        |                                   |
        +-----------------+-----------------+
                 |                       |
       +------------------+    +------------------+
       | Database Server  |    |   Redis Server   |
       +------------------+    +------------------+

Local Emulation with Aspire and Service Discovery

Example Web project on GitHub
Example Aspire project on GitHub

To emulate the production environment locally, Fluent CMS leverages Aspire. Here's an example setup:

var builder = DistributedApplication.CreateBuilder(args);

// Adding Redis and PostgreSQL services
var redis = builder.AddRedis(name: CmsConstants.Redis);
var db = builder.AddPostgres(CmsConstants.Postgres);

// Configuring the web project with replicas and references
builder.AddProject<Projects.FluentCMS_Blog>(name: "web")
    .WithEnvironment(CmsConstants.DatabaseProvider, CmsConstants.Postgres)
    .WithReference(redis)
    .WithReference(db)
    .WithReplicas(2);

builder.Build().Run();

Benefits:

• Simplified Configuration:
  No need to manually specify endpoints for the database or Redis servers. Configuration values can be retrieved using:

  builder.Configuration.GetValue<string>();
  builder.Configuration.GetConnectionString();
  

• Realistic Testing:
  The local environment mirrors the production architecture, ensuring seamless transitions during deployment.

By adopting these caching and deployment strategies, Fluent CMS ensures improved performance, scalability, and ease of configuration.

Query with Document DB

Optimizing query performance by syncing relational data to a document database, such as MongoDB, significantly improves speed and scalability for high-demand applications.

Limitations of ASP.NET Core Output Caching

ASP.NET Core's output caching reduces database access when repeated queries are performed. However, its effectiveness is limited when dealing with numerous distinct queries:

• The application server consumes excessive memory to cache data. The same list might be cached multiple times in different orders.
• The database server experiences high load when processing numerous distinct queries simultaneously.

Using Document Databases to Improve Query Performance

For the query below, FluentCMS joins the post, tag, category, and author tables:

query post_sync($id: Int) {
  postList(idSet: [$id], sort: id) {
    id, title, body, abstract
    tag {
      id, name
    }
    category {
      id, name
    }
    author {
      id, name
    }
  }
}

By saving each post along with its related data as a single document in a document database, such as MongoDB, several improvements are achieved:

• Reduced database server load since data retrieval from multiple tables is eliminated.
• Reduced application server processing, as merging data is no longer necessary.

Performance Testing

Using K6 scripts with 1,000 virtual users concurrently accessing the query API, the performance difference between PostgreSQL and MongoDB was tested, showing MongoDB to be significantly faster:

export default function () {
    const id = Math.floor(Math.random() * 1000000) + 1; // Random id between 1 and 1,000,000
    /* PostgreSQL */
    // const url = `http://localhost:5091/api/queries/post_sync/?id=${id}`;

    /* MongoDB */
    const url = `http://localhost:5091/api/queries/post/?id=${id}`;

    const res = http.get(url);

    check(res, {
        'is status 200': (r) => r.status === 200,
        'response time is < 200ms': (r) => r.timings.duration < 200,
    });
}
/*
MongoDB:
     http_req_waiting...............: avg=50.8ms   min=774µs    med=24.01ms max=3.23s    p(90)=125.65ms p(95)=211.32ms
PostgreSQL:
     http_req_waiting...............: avg=5.54s   min=11.61ms med=4.08s max=44.67s  p(90)=13.45s  p(95)=16.53s
*/

Synchronizing Query Data to Document DB

Architecture Overview

Enabling Message Publishing in WebApp

To enable publishing messages to the Message Broker, use Aspire to add a NATS resource. Detailed documentation is available in Microsoft Docs.

Add the following line to the Aspire HostApp project:

builder.AddNatsClient(AppConstants.Nats);

Add the following lines to the WebApp project:

builder.AddNatsClient(AppConstants.Nats);
var entities = builder.Configuration.GetRequiredSection("TrackingEntities").Get<string[]>()!;
builder.Services.AddNatsMessageProducer(entities);

FluentCMS publishes events for changes made to entities listed in appsettings.json:

{
  "TrackingEntities": [
    "post"
  ]
}

Enabling Message Consumption in Worker App

Add the following to the Worker App:

var builder = Host.CreateApplicationBuilder(args);

builder.AddNatsClient(AppConstants.Nats);
builder.AddMongoDBClient(AppConstants.MongoCms);

var apiLinksArray = builder.Configuration.GetRequiredSection("ApiLinksArray").Get<ApiLinks[]>()!;
builder.Services.AddNatsMongoLink(apiLinksArray);

Define the ApiLinksArray in appsettings.json to specify entity changes and the corresponding query API:

{
  "ApiLinksArray": [
    {
      "Entity": "post",
      "Api": "http://localhost:5001/api/queries/post_sync",
      "Collection": "post",
      "PrimaryKey": "id"
    }
  ]
}

When changes occur to the post entity, the Worker Service calls the query API to retrieve aggregated data and saves it as a document.

Migrating Query Data to Document DB

After adding a new entry to ApiLinksArray, the Worker App will perform a migration from the start to populate the Document DB.

Replacing Queries with Document DB

Architecture Overview

To enable MongoDB queries in your WebApp, use the Aspire MongoDB integration. Details are available in Microsoft Docs.

Add the following code to your WebApp:

builder.AddMongoDBClient(connectionName: AppConstants.MongoCms);
var queryLinksArray = builder.Configuration.GetRequiredSection("QueryLinksArray").Get<QueryLinks[]>()!;
builder.Services.AddMongoDbQuery(queryLinksArray);

Define QueryLinksArray in appsettings.json to specify MongoDB queries:

{
  "QueryLinksArray": [
    { "Query": "post", "Collection": "post" },
    { "Query": "post_test_mongo", "Collection": "post" }
  ]
}

The WebApp will now query MongoDB directly for the specified collections.

Integrating it into Your Project

Follow these steps to integrate Fluent CMS into your project using a NuGet package.

• Create a New ASP.NET Core Web Application.

• Add the FluentCMS NuGet Package: To add Fluent CMS, run the following command:

  dotnet add package FluentCMS
  

• Modify Program.cs: Add the following line before builder.Build() to configure the database connection (use your actual connection string):

  builder.AddSqliteCms("Data Source=cms.db");
  var app = builder.Build();
  

  Currently, Fluent CMS supports AddSqliteCms, AddSqlServerCms, and AddPostgresCms.

• Initialize Fluent CMS: Add this line after builder.Build() to initialize the CMS:

  await app.UseCmsAsync();
  

  This will bootstrap the router and initialize the Fluent CMS schema table.

• Optional: Set Up User Authorization: If you wish to manage user authorization, you can add the following code. If you're handling authorization yourself or don’t need it, you can skip this step.

  builder.Services.AddDbContext<AppDbContext>(options => options.UseSqlite(connectionString));
  builder.AddCmsAuth<IdentityUser, IdentityRole, AppDbContext>();
  

  If you'd like to create a default user, add this after app.Build():

  InvalidParamExceptionFactory.CheckResult(await app.EnsureCmsUser("[e-mail]", "Admin1!", [Roles.Sa]));
  

Once your web server is running, you can access the Admin Panel at /admin and the Schema Builder at /schema.

You can find an example project here.

Adding Business Logic

Learn how to customize your application by adding validation logic, hook functions, and producing events to Kafka.

Adding Validation Logic with Simple C# Expressions

Simple C# Validation

You can define simple C# expressions in the Validation Rule of attributes using Dynamic Expresso. For example, a rule like name != null ensures the name attribute is not null.

Additionally, you can specify a Validation Error Message to provide users with feedback when validation fails.

Using Regular Expressions

Dynamic Expresso supports regular expressions, allowing you to write rules like Regex.IsMatch(email, "^[a-zA-Z0-9_.+-]+@[a-zA-Z0-9-]+\\.[a-zA-Z0-9-.]+$").

Note: Since Dynamic Expresso doesn't support verbatim strings, you must escape backslashes (\).

Extending Functionality with Hook Functions

To implement custom business logic, such as verifying that a teacher entity has valid email and phone details, you can register hook functions to run before adding or updating records:

var registry = app.GetHookRegistry();

// Hook function for pre-add validation
registry.EntityPreAdd.Register("teacher", args =>
{
    VerifyTeacher(args.RefRecord);
    return args;
});

// Hook function for pre-update validation
registry.EntityPreUpdate.Register("teacher", args =>
{
    VerifyTeacher(args.RefRecord);
    return args;
});

Producing Events to an Event Broker (e.g., Kafka)

To enable asynchronous business logic through an event broker like Kafka, you can produce events using hook functions. This feature requires just a few additional setup steps:

• Add the Kafka producer configuration:

  builder.AddKafkaMessageProducer("localhost:9092");
  

• Register the message producer hook:

  app.RegisterMessageProducerHook();
  

Here’s a complete example:

builder.AddSqliteCms("Data Source=cmsapp.db").PrintVersion();
builder.AddKafkaMessageProducer("localhost:9092");
var app = builder.Build();
await app.UseCmsAsync(false);
app.RegisterMessageProducerHook();

With this setup, events are produced to Kafka, allowing consumers to process business logic asynchronously.

Development Guide

The backend is written in ASP.NET Core, the Admin Panel uses React, and the Schema Builder is developed with jQuery.

System Overviews

• Backend Server
• Admin Panel UI
• Schema Builder

Backend Server

• Tools:
  • ASP.NET Core
  • SqlKata: SqlKata

Admin Panel UI

• Tools:
  • React
  • PrimeReact: PrimeReact UI Library
  • SWR: Data Fetching/State Management

Schema Builder UI

• Tools:
  • jsoneditor: JSON Editor

Testing Strategy

This chapter describes Fluent CMS's automated testing strategy

Fluent CMS favors integration testing over unit testing because integration tests can catch more real-world issues. For example, when inserting a record into the database, multiple modules are involved:

• EntitiesController
• EntitiesService
• Entity (in the query builder)
• Query executors (e.g., SqlLite, Postgres, SqlServer)

Writing unit tests for each individual function and mocking its upstream and downstream services can be tedious. Instead, Fluent CMS focuses on checking the input and output of RESTful API endpoints in its integration tests.

However, certain cases, such as the Hook Registry or application bootstrap, are simpler to cover with unit tests.

Unit Testing /fluent-cms/server/FluentCMS.Test

This project focuses on testing specific modules, such as:

• Hook Registry
• Application Bootstrap

Integration Testing for FluentCMS.Blog /fluent-cms/server/FluentCMS.Blog.Tests

This project focuses on verifying the functionalities of the FluentCMS.Blog example project.

New Feature Testing /fluent-cms/server/FluentCMS.App.Tests

This project is dedicated to testing experimental functionalities, like MongoDB and Kafka plugins.