Counting since 20-10-2023

URDF Analyzer

This is a tool developed to analyze URDF Collections. It is created to be used as a standalone tool or together with the URDF Dataset. It allows analyzing single URDF files, or directories containing URDF files where schemas of the results can be generated.

Command Line Interface

The following shows the different use cases that the urdf_analyzer supports.

Generate schemas

Obtain the model-information of the urdf files found in a specific directory (recursively).

urdf_analyzer generate-schemas model-info --urdf-search-dir <directory-to-search-for-urdfs>

Compare parsing performance of urdf tools on urdf files found in a specific directory (recursively).

urdf_analyzer generate-schemas tool-cmp --urdf-search-dir <directory-to-search-for-urdfs>

Compare duplicates in a specified folder.

urdf_analyzer generate-schemas duplicates-cmp --duplicates-dir <directory-containing-duplicates>

Todo tool:

• Testing
  • create tests for model_analysis
  • create tests for joint
  • create tests for link
  • create tests for urdf_parser
  • add testing of the 'full' argument in the cli tests
  • add testing of get_parsing_information in the api tests
  • add testing of parsing_information in the cli tests
  • add a test to check that the generated schema contains all of the necessary information, e.g. filenames..
• Development API
  • in the save information func, check that the checking of existence of the output file is done correctly with regards to the directory
  • figure out how to export the information, csv? txt? latex table?
  • add analysis of urdf using the urdf parsers
  • does it make sense to add an exclusion of files or subfolders when performing the urdf-search-dir?
  • add domain knowledge, e.g. the user should be able to specify the system is a robotic arm with X DOF, and then the URDF analyser can analyse it and check that this is correct
  • check if it would be better/faster to load one URDF loader, and then run through all the files, or if the current method is ok.
  • fix the checking of 'xacro' and 'package' to ensure it's not in the comments of the xml file
  • create a get_mesh_analysis_schema function that just takes out the mesh values from the model_information dataframe
  • consider adding a textual description of the differences between the duplicate urdfs
  • setup the check_urdf as a parser, and make that the default one
• Development CLI
  • add cli for geometry (.dae, .stl, .obj)
  • add cli for file analysis ('xacro' in file, 'package' in file)
• Documentation
  • describe the different schemas that can be generated

Notes:

To setup the package in development, run python setup.py develop from the root directory.