InfoBIM

InfoBIM is the OpenBIM command line of the OntoBDC ecosystem. It turns an ordinary folder of building information into an InfoBIM project: IFC models, DXF/DWG drawings, PDFs, images and spreadsheets are described in a semantic graph. The files stay where they are, in their own formats.

InfoBIM is a thin domain layer on top of OntoBDC, the generic semantic runtime:

• OntoBDC owns containers, datasets, the storage index, the RO-Crate and Data Package manifests, health checks and state machines.
• InfoBIM adds only what depends on IFC and buildingSMART semantics.

The ontologies both use, such as AECO kinds, file encodings and facades, come from the brasidatacenter package. They are looked up by the IRI they are published at.

The same CLI runs in the browser in databim.tech, through Pyodide.

Install

InfoBIM needs Python 3.11 or later, which brasidatacenter requires. Install its two sibling packages first, then InfoBIM:

pip install brasidatacenter ontobdc infobim

To work on the code, install the three repositories in editable mode in one virtual environment (reinstall.sh does it for InfoBIM):

pip install -e ../brasidatacenter -e ../ontobdc-wip -e .

Optional extras:

To read DWG drawings, install the ODA File Converter. InfoBIM finds it on the PATH, or at the executable named by INFOBIM_ODA_FILE_CONVERTER.

Quickstart

mkdir ~/obras && cd ~/obras
infobim init                              # makes this folder a storage root

infobim project --create "Hospital Norte"   # creates and registers ./hospital-norte
cd hospital-norte

cp ~/Downloads/estrutura.ifc ~/Downloads/planta.dxf .
infobim project --refresh                 # registers the new files in the project

infobim project --inspect                 # what the project holds, from its IfcProject down
infobim project --entity                  # the entities the project holds, by dataset
infobim project --health                  # every check the project must pass
infobim project --list                    # (from anywhere under the root) every project

Every command answers in JSON with --json. This is how databim.tech reads the answers.

Concepts

Storage root

infobim init makes a folder a storage root by creating a .__ontobdc__/ directory in it:

• config.yaml is the marker that identifies the root.
• storage.ttl is the index of the projects registered under the root.

Commands find the root by walking up from the current directory to that marker. A root can be moved, or created in one file system and opened in another (for example, created in the browser and opened on disk): the locations recorded in the index are read relative to where the root is now. A recorded location outside the root is an error, not something silently tolerated.

Project

A project is an OntoBDC container, a folder registered in the root's index, that carries InfoBIM's reserved dataset, .__infobim__/. That dataset holds the IfcProject declaration. The IfcProject's GlobalId is the project's identity.

hospital-norte/
├── .__ontobdc__/
│   ├── container.ttl             the container's metadata and its datasets
│   ├── ro-crate-metadata.json    every file of the project, with its media type
│   └── datapackage.json          the tabular files, for frictionless
├── .__infobim__/                 the reserved InfoBIM dataset
│   ├── .__ontobdc__/dataset.ttl
│   └── payload/
│       ├── triple/ifc_project.ttl    the IfcProject (GlobalId, name)
│       └── linkset/dataset_facade.ttl
├── estrutura.ifc
└── planta.dxf

--create <name> takes a name. The project goes into a subfolder named after its slug ("Hospital Norte" → hospital-norte/) under the current storage root. Creating a project also refreshes it, so a new project already passes project --health.

Choosing the project a command acts on

Every project command acts on:

• the project of the current directory, or
• the project whose IfcProject GlobalId is given with --global-id <GlobalId>.

Refresh

infobim project --refresh brings a project in line with its folder, in two stages:

1. OntoBDC's container refresh. It checks the container and each of its datasets, repairing what fails. It removes stray files. It rewrites the Data Package and the RO-Crate, so files added, changed or removed are recorded. A dataset still unhealthy after its repair fails the refresh, naming the dataset and the checks it fails.
2. InfoBIM's project refresh. It reconciles the reserved dataset and the IfcProject on top of the refreshed container.

Each file is recorded in the RO-Crate with a media type (encodingFormat):

• For the domain's own formats, it is the one the AECO ontology declares: application/x-step for IFC, image/vnd.dxf for DXF.
• For any other file, it is the type built into Python.
• An extension that neither knows gets no media type.

The result is the same on every machine.

Commands

Every project command also accepts --global-id <GlobalId> instead of the current directory, and --json for a machine-readable answer. infobim --help lists every command with its flags.

infobim ifc --point opens the drawing in the 2D viewer to pick the points. It creates the elements in the model --ifc-model-path names. Without it, they go into the project's own model, named after its IfcProject GlobalId; that model is created and declared in the project if it does not exist yet.

Source layout

src/infobim/
├── cli/        entry point, init, --version, the welcome screen
├── project/    project lifecycle: create, list, inspect, health, refresh, update, attach, delete
├── ifc/        IFC elements: creation from points, IFC model bootstrap
├── drawing/    drawings: DXF reading, DWG → DXF through the ODA File Converter
├── context/    the ontology dictionary command
├── 2d/, _2d/   the 2D viewer (optional PySide6 extra) and its proxy
├── _3d/        the 3D viewer (optional PySide6 extra)
├── dev/        proxy to the ontobdc-dev tools
└── shared/     configuration shared by the components
legacy/         earlier implementation and its test suite, kept for reference; not packaged

Each component follows OntoBDC's plugin layout: plugin/command/, plugin/parameter/, plugin/check/ (with a check.py and its hotfix.py), plugin/capability/, plugin/machine/ (state machines in YAML), adapter/ and domain/. The CLI discovers commands from these folders.

Tests

Tests run through the ontobdc-dev tools:

infobim dev test          # the InfoBIM test suite
infobim dev test --e2e    # the end-to-end tests, which run the real infobim executable

The previous end-to-end suite now lives under legacy/test/ with the earlier implementation. The active suite under test/ is being rebuilt.

License

Apache License 2.0.