CBIMS.IFCNormalization is a tool to "normalize" an IFC file (in ISO 10303-21). The normalized IFC file can be used for better comparison of changes between versions, and for storage in Git-like tools.

The tool has the following features:

• Merging redundant nodes in an IFC file.
• Re-assigning stable integer row IDs according to the row contents and references.
• Sorting the rows, and (optionally) segmenting the rows into chunks.

Please refer to:

Liu H, Gao G, Gu M. A Parallel IFC Normalization Algorithm for Incremental Storage and Version Control. International Workshop on Intelligent Computing in Engineering (EG-ICE), 2023: 511-520.

https://arxiv.org/ftp/arxiv/papers/2312/2312.14931.pdf

Patent (China):

刘寒, 高歌, 顾明. 一种图数据处理与增量存储方法、装置及计算机设备. 中国专利: ZL 202211625638.5 . 2022年12月16日.

The project uses the following dependencies from the NuGet:

• Xbim.IO.MemoryModel - https://github.com/xBimTeam/XbimEssentials

Other resources contained in the repository:

• CBIMS.IFCNormalization.Core/ThirdParty/IfcGuid.cs - Methods for converting between GUIDs and IFC GlobalIds, available from: https://github.com/hakonhc/IfcGuid/blob/master/IfcGuid/IfcGuid.cs .

• CBIMS.IFCNormalization.CMD - The command-line entrance of the tool.
• CBIMS.IFCNormalization.Core - The core implementation of the IFC normalization algorithm.
• CBIMS.IFCNormalization.Interface - The interface for accessing the input IFC data structure.
• CBIMS.IFCNormalization.Xbim - The implementation of the interface for getting IFC data using Xbim

dotnet CBIMS.IFCNormalization.CMD.dll -i <input path>
 [-o <output path>]
 [--level <chunk level>]
 [--spare <chunk spare rate>]
 [--parallel <true|false>]
 [--exp_chunk_num <true|false>]
 [--rm_ownerhistory <true|false>]
 [--do_segment <false|true>]
 [--just_compress <false|true>]

-i: An input IFC path.

-o: <default {inputPath}/{inputFileName}.norm.ifc> The output IFC path.

--level: <default 5> The level of chunk size.

Level Chunk size Max chunk numbers
3 10000000 214
4 1000000 2147
5 100000 21474
6 10000 214748
7 1000 2147483
8 100 21474836
9 10 214748364

--spare: <default 2.0> A real spare rate greater than 1.0 to make more efficient assignment of node storage.

--parallel: <default true> Use multi-core CPU to speed up calculation.

--exp_chunk_num: <default true> Using the exponential function for increasing the number of chunks of each type.

--rm_ownerhistory: <default true> Removing IfcOwnerHistory references for all IfcRoot nodes on output.

--do_segment: <default false> Adding "/*========*/" as segmentation for each chunk.

--just_compress: <default false> I do not want to re-organize the IDs but just want to merge the redundant nodes.
* Ignoring most of the parameters above except '--parallel' and '--rm_ownerhistory'.
* The default output file name will be '{inputFileName}.compress.ifc'.

CBIMS.IFCNormalization uses GNU Lesser General Public License (LGPL). Please refer to: https://www.gnu.org/licenses/lgpl-3.0.en.html

CBIMS (Computable Building Information Modeling Standards) is a set of research projects aiming at the development of computable BIM standards for the interoperability of BIM data exchange and the automation of BIM-based workflow. The projects are led by the BIM research group at the School of Software, Tsinghua University, China.