Document validation and header extraction flow

[aothms]

flowchart LR
    %% Define overall subgraph for clarity
    subgraph ValidationRequest

        A((start))
        A --> B        
        
       subgraph Serial Chain
            B(syntax_validation)
            C(parse_info)
            D("gherkin (blockers)")

            B -->|Valid| C
            C -->|Valid| D
        end
        
        B -->|Invalid| stop1((Stop))
        C -->|Invalid| stop2((Stop))
        D -->|Invalid| stop3((Stop))

        subgraph Parallel Group
            E(schema_validation)
            %%F(bsdd_validation)
            G["normative_rules"]
            H["gherkin_ia"]
            I["gherkin_ip"]
            J(industry_practices)
        end
        
        D --> E
        %%D --> F
        D --> G
        G --> H
        G --> I
        D --> J
    
    end

Loading

[Ghesselink]

flowchart LR
    %% Define overall subgraph for clarity
    subgraph ValidationRequest

        A((Start))
        A --> B(syntax_validation)

        B -->|Valid| C(parse_info)
        B -->|Invalid| stopSyntax((Stop))

        subgraph Serial Chain
            C --> C1{Invalid_Schema?}

            C1 -- Yes --> HVAL_SPF[(HeaderV_SPF)]

            C1 -- No --> HVAL_IFC[(HeaderV_IfcOpenShell)]

            HVAL_SPF --> HI[(File_Info)]
            HVAL_SPF --> stopHeaderSPF((Stop))

            HVAL_IFC --> HI
            HVAL_IFC --> D("gherkin (blockers)")
        end

        subgraph Parallel GroupF
            E(schema_validation)
            %%F(bsdd_validation)
            G["normative_rules"]
            H["gherkin_ia"]
            I["gherkin_ip"]
            J(industry_practices)
        end
        
        D -->|Valid| E
        %%D -->|Valid| F
        D -->|Valid| G
        D -->|Invalid| stopGherkin((Stop))
        G --> H
        G --> I
        D -->|Valid| J
    
    end

Loading

[aothms]

Invalid_Schema?

Is not the most accurate I guess. It's essentially: "schema supported by ifcopenshell". Later we have "gherkin blockers", which is at this point in time only IFC101 which checks whether the schema is 'actual' ('IFC4X3_ADD2' or 'IFC4' or 'IFC2X3').

---

But thanks. This is 100% clear and shows why we don't have File_Info data when syntax_validation reports Invalid. 👍

[Ghesselink]

Is not the most accurate I guess. It's essentially: "schema supported by ifcopenshell"

That's true, I had to abbreviate the terms a little but this corner was then cut a bit too short.

I've also made a sketch for IVS-383;

• Always run SPF-based header validation
• If the syntax error gives a 'data error', still run the header validation (but stop afterwards). This way, we can link vendors with syntax errors. If the syntax error is in the header, stop.
• Add the instance_completion subtask. We use this task for the report (show failing id and entity type)

flowchart LR
    subgraph ValidationRequest

        A((Start))
        A --> B(syntax_validation)
        B -->|Data_Error or Valid| C(parse_info)
        C --> HVAL_SPF[(HeaderV_SPF)]
        B -->|Header_Error| stopSyntax((Stop))

        subgraph Parse Info
            HVAL_SPF --> | Store| HI[(File_Info)]
            HVAL_SPF -->|Valid| D("gherkin (blockers)")
            HVAL_SPF -->|Invalid or Error| stopHeader((Stop))
        end

        subgraph Parallel Group
            E(schema_validation)
            %%F(bsdd_validation)
            G["normative_rules"]
            H["gherkin_ia"]
            I["gherkin_ip"]
            J(industry_practices)
        end
        
        D -->|Valid| E
        %%D -->|Valid| F
        D -->|Valid| G
        D -->|Invalid| stopGherkin((Stop))
        G --> H
        G --> I
        D -->|Valid| J

        E --> K(instance_completion)
        G --> K
        H --> K
        I --> K
        J --> K

    end

Loading

[aothms]

Really nice. Just a note for myself, I'll elaborate on this later. I think I'd propose two syntax validation task. syntax_validation_header and syntax_validation_data. Because failing syntax_validation_data should probably still be a blocker for any of the other gherkin rules.

[Ghesselink]

Updated flowchart;

• Syntax_validation task split up in header and data component
• Header_validation_task instead of parse_info.
• Header validation always uses SPF, not depending on (absence of) IFC schema errors
• First c++/ifcopenshell.open() pre-check before first syntax check

flowchart TD

    %% Tasks
    A([**Start**])
    B1["**cpp_open_errors**<br/><i>ifcopenshell.open(fn)</i><br/><i>➤ Syntax</i>"]
    B2["**syntax_validation_data**<br/><i>simple_spf (data-only)</i><br/><i>➤ Syntax</i>"]
    B3["**syntax_validation_header**<br/><i>ifcopenshell.simple_spf</i><br/><i>(header-only API)</i><br/><i>check for errors</i><br/><i>➤ Syntax</i>"]
    C["**header_validation**<br/><i>ifcopenshell.simple_spf</i><br/><i>(header-only API)</i><br/><i>parse information</i><br/><i>➤ File Information</i>"]
    D["**gherkin_blockers**<br/><i>Blocker rules</i><br/><i></i>"]
    F["**instance_completion**"]
    END([**Validation Complete**])

    %% Stops 
    STOP_CPP((stop))
    STOP_SYNTAX_DATA((stop))
    %% STOP_HEADER((stop))
    STOP_GHERKIN((stop))

    %% Grouped Parallel Tasks
    subgraph Parallel_Group [ ]
        direction TB
        E1["**schema_validation**<br/><i>ifcopenshell.validate</i><br/><i>EXPRESS schema</i><br/><i>compliance</i><br/><i>➤ IFC Schema</i>"]
        E2["**normative_rules**"]
        E3["**gherkin_ia**<br/><i>Implementer Agreement</i><br/><i>rules</i><br/><i>➤ Normative IFC rules</i>"]
        E4["**gherkin_ip**<br/><i>Information Proposition</i><br/><i>rules</i><br/><i>➤ Normative IFC rules</i>"]
        E5["**industry_practices**<br/><i>Best practices / Warnings</i><br/><i>rules</i><br/><i>➤ Industry Practices</i>"]
    end

    %% Connections
    A --> B1
    B1 -->|Valid| B2
    B1 -->|Error| STOP_CPP

    B2 -->|Valid| B3
    B2 -->|Error| STOP_SYNTAX_DATA

    B3 --> C

    C -->|Valid| D
    %% C -->|Invalid| STOP_HEADER

    D -->|Valid| E1
    D -->|Valid| E2
    D -->|Valid| E5
    D -->|Invalid| STOP_GHERKIN

    E2 --> E3
    E2 --> E4

    E1 --> F
    E3 --> F
    E4 --> F
    E5 --> F

    F --> END

Loading

[Ghesselink]

flowchart TD

    %% Main Steps
    A([**Start**])
    B2["**syntax_validation_header**<br/>Uses: simple_spf (header-only)<br/>Displayed in: Syntax"]
    B3["**header_validation**<br/>Gathers file information from header<br/>"]
    C["**syntax_validation_data**<br/>Uses: simple_spf (data-specific)<br/>Displayed in: Syntax"]
    D["**gherkin_blocker**<br/>Executes: IFC101<br/>Displayed in: Schema"]
    F["**instance_completion**<br/>Link errors to IFC entities"]
    END([**Validation Complete**<br/>Send resuls to<br/>frontend or API])

    %% Stops (smaller and rounded)
    STOP_SYNTAX_DATA((stop))
    STOP_HEADER((stop))
    STOP_GHERKIN((stop))

    %% Grouped Parallel Tasks
    subgraph Parallel_Group [ ]
        direction TB
        E3["**digital signatures**<br/>Displayed: Next to filename"]
        E1["**schema_validation**"]
        E2["**normative_rules**"]
        E5["**industry_practices**"]
    end

    %% Connections
    A --> B2

    B2 -->|Valid| B3
    B2 -->|Error| STOP_SYNTAX_DATA

    B3 --> C

    C -->|Valid| D
    C -->|Invalid| STOP_HEADER

    D -->|Valid| E1
    D -->|Valid| E3
    D -->|Valid| E2
    D -->|Valid| E5
    D -->|Invalid| STOP_GHERKIN


    E1 --> F
    E2 --> F
    E5 --> F

    F --> END

Loading