Created: 2023-10-21 | Updated: 2024-09-21

UML ingestion from mermaid/plantuml formats

New (v0.3.15)

Table of contents

  1. UML ingestion from mermaid/plantuml formats
  2. Source infrastructure Diagram:
  3. Mermaid UML Diagrams
    1. Mermaid UML rendered diagram:
    2. Advanced for Geeks(Mermaid UML sources):
    3. Mermaid UML rendered diagram:
    4. Advanced for Geeks(Mermaid UML sources):
  4. Ingesting UML metadata into multicloud-diagrams:
    1. Load initial drawio infra diagram
    2. Define mappings
    3. Mapping format structure
    4. Generate MetaInformation Layer
    5. Result File
    6. Fully Editable Diagram with Layers from UML

Record are rows of text.

Latest Multicloud-diagrams feature allows automatically parse sequence UML diagrams, detect actors, actions and represent them on source Diagram as additional connected nodes with aliases. This makes documenting infrastructure even more robust. Framework automatically detects diagram type passed into invocation parameter whether it is plantuml or mermaid.

Source infrastructure Diagram:

We have existing infrastructure diagram that describes AWS components of architecture:

Starting

Mermaid UML Diagrams

Following UML diagram in mermaid format describes File upload stage:

Mermaid UML rendered diagram:

sequenceDiagram
    actor Donald
    participant S3Bucket
    participant LambdaFunction1
    participant SNS
    participant SQS

    Donald->>S3Bucket: Put new file into S3 bucket
    S3Bucket->>LambdaFunction1: LifeConfig Rule invokes lambda function
    LambdaFunction1->>SNS: Send Event to SNS that file processing is finished
    SNS->>SQS: Fanout Event to SQS

Advanced for Geeks(Mermaid UML sources):

sequenceDiagram
    actor Donald
    participant S3Bucket
    participant LambdaFunction1
    participant SNS
    participant SQS

    Donald->>S3Bucket: Put new file into S3 bucket
    S3Bucket->>LambdaFunction1: LifeConfig Rule invokes lambda function
    LambdaFunction1->>SNS: Send Event to SNS that file processing is finished
    SNS->>SQS: Fanout Event to SQS

Next UML diagram in plantUML format that describes next Processing stage:

Mermaid UML rendered diagram:

sequenceDiagram
    participant SQS
    participant LambdaFunction2
    participant Storage

    SQS->>LambdaFunction2: poll message from SQS and invoke consumer lambda function
    LambdaFunction2->>Storage: Update DynamoDB Table. Save image status.

Advanced for Geeks(Mermaid UML sources):

@startuml
    queue SQS
    participant LambdaFunction2
    database Storage

    SQS->>LambdaFunction2: poll message from SQS and invoke consumer lambda function
    LambdaFunction2->>Storage: Update DynamoDB Table. Save image status.
@enduml

Ingesting UML metadata into multicloud-diagrams:

Load initial drawio infra diagram

As usual, we create a new diagram using Diagrams-As-a-Code or open exising one:

from multicloud_diagrams import MultiCloudDiagrams
from samples.samples.aws_service_end_2_end import prepare_end2end


def main():
    mcd = MultiCloudDiagrams()

    prefix = 'prod'
    prev_file = f'../output/output.{prefix}_plantuml.end2end.drawio'

    mcd.read_coords_from_file(prev_file)

Define mappings

Now it is time to define mappings between diagram elements IDs and aliases used in UML sequence diagram.

    mcd.read_uml_mappings('uml_mapping.yml')

Mapping format structure

Mappings are defined in yaml format, they specify relations between actor from UML and node_id from diagrams.

nodes:
  - actor: S3Bucket
    node_id: s3:arn:aws:s3:::bucket_name/key_name
  - actor: LambdaFunction1
    node_id: lambda_function:arn:aws:lambda:eu-west-1:123456789012:function:producer-lambda
  - actor: LambdaFunction2
    node_id: lambda_function:arn:aws:lambda:eu-west-1:123456789012:function:consumer-lambda
  - actor: SNS
    node_id: sns:arn:aws:sns:eu-west-1:123456789012:internal.fifo
  - actor: SQS
    node_id: sqs:arn:aws:sqs:eu-west-1:123456789012:int-eu-live-events.fifo
  - actor: Storage
    node_id: dynamo:arn:aws:dynamodb:eu-west-1:123456789012:table/table1

Generate MetaInformation Layer

For each UML diagram, multicloud-diagrams will created Layer in draw-io and augment relations and actions to existing elements.

Customization of colors, fonts, arrows are fully supported. In this example we are creating distinct styles for each UML diagram (they will be rendered with different colors).

    label_style = {
        'labelBackgroundColor': 'none',
        'fontColor': '#FF3333',
        'fontStyle': 0,
        'fontSize': 14
    }
    style1 = {
        'orthogonalLoop': '1',
        'edgeStyle': 'orthogonalEdgeStyle',
        'curved': '1',
        'startArrow': 'oval',
        'endArrow': 'classicThin',
        'dashed': '1',
        'strokeColor': '#FF3333',
        'strokeWidth': '3',
        'fontSize': '22'
    }
    mcd.read_uml_from_file('file_upload.mermaid', edge_style=style1, label_style=label_style)

Augment second UML diagram into multicloud-diagrams with different styles instrumented.

    style2 = {
        'orthogonalLoop': '1',
        'edgeStyle': 'orthogonalEdgeStyle',
        'curved': '1',
        'startArrow': 'oval',
        'endArrow': 'classicThin',
        'dashed': '1',
        'strokeColor': '#0000FF',
        'strokeWidth': '3',
        'fontSize': '22'
    }

    label_style = {
        'labelBackgroundColor': 'none',
        'fontColor': '#0000FF',
        'fontStyle': 0,
        'fontSize': 14
    }
    mcd.read_uml_from_file('process.mermaid', edge_style=style2, label_style=label_style)

There is option to overwrite the same file or produce a new one. We will use the second option.

    result_file = f'../output/output.{prefix}_uml.end2end.drawio'
    mcd.export_to_file(result_file)

Result File

result

Fully Editable Diagram with Layers from UML

Each layer has a name of UML file that we ingested, all actions from each UML are marked with numbers, customized with colors and added on top of existing infra elements (draggable, editable):

drawio-uml.gif