Skip to content
tools to process actions on revit models from cli
Branch: master
Clone or download
hdm-dt-fb fix log path
Latest commit f95d89a Jun 24, 2019


python micro framework to process actions on revit models from cli/command line

how it works:

  • you initialize it with a command to specifying the task, revit model and an optional timeout.
  • will spin off a subprocess to write a journal file and addin according to your specified task and project.
  • it will then run Revit according to this journal file, opening a detached version of your model and run the specified action like a revitpythonshell script.
  • if the journal file cannot be run to completion the subprocess is killed and an error is logged. the logging journal file will be parsed and a notify email will inform you in case the model is corrupt.

it requires/is currently run on:

  • cpython (64bit recommended) >= 3.6 (with additional modules)
    to install modules on your machine run
    pip install beautifulsoup4 bokeh colorama colorful docopt numpy pandas psutil olefile slackclient rvt_detector rjm
  • Autodesk Revit® (currently tested on versions 2015.7, 2016.2, 2017.2)
  • see install_guide for help/further information.

how to get started:

  • when the above mentioned requirements are met and this repo is cloned to your preferred path lets get started with a common task (read qc stats (qc meaning "quality check" in this context)) in a few steps:

  • run a qc read out from cli
    run from command line: now we can run the qc_model without even touching Revit and get interactive html graphs produced. Compose a command line instruction consisting of the following:

    your CPython interpreter
    your path to
    command type
    project name
    full path to the project Revit model
    a timeout for the process (optional - default is 60 seconds)

    so it could look like this:

    "C:\Program Files\Python36\python.exe"
    --timeout 600

    Just concatenate it (put it into one line):
    "C:\Program Files\Python36\python.exe" D:/testrun/934_rvt_model_services/ qc 123_N D:/testmodel/123_N.rvt --timeout 600
    Open a command line ("Win > type 'cmd'") paste it in("right-click > paste") and run it.
    If you want to write the html to another directory you can use the optional switch "--html_path" followed by a path.
    Here is how this looks on my screen:


  • step 2: let task scheduler repeat your task
    for recurring tasks hook it up to Windows® task scheduler: Open Task scheduler and create a new basic task
    - give it a name
    - set your interval rate e.g. daily
    - set your start time
    - action: start program:
    "C:\Program Files\Python36\python.exe"
    Add Arguments:
    D:/testrun/934_rvt_model_services/ qc 123_N D:/testmodel/123_N.rvt --timeout 600
    - Finish and test if it works: "Right-Click > Run"

further options:

  • If you are too lazy to type the path for different projects all the time (like me) you can set them up in a config.ini like this:

    path = D:/testmodel/123_N.rvt

    With this setup you can substitue the actual full_model_path in your command with just: cfg

  • If you want to get notified on model corruption or circular links just add the flag --notify to setup notifications (currently email or slack) refer to the of the notify module.

typical use cases(recurring tasks run via schedule Task Scheduler):

  • extraction of qc data to be gathered in csv table and visualized on interactive html graphs using bokeh
  • check on model corruption with audit canary
  • export of DWF, DWG, PDF or IFC (so far only DWF export implemented)
  • export model warnings (API-less journal file warnings export)

currently implemented tasks:

  • qc: rvt model qc statistics on workshared models


  • dwf: dwf sheet exports of sheet set "Auto_PDF_DWF" on workshared models

  • warnings: model warnings export on workshared models


  • pulse: bokeh graph showing the job log graphically(which project was process in what time, did it timeout -> red bar).
    run separately from process_model with: "python" from commands/pulse/ directory.


  • audit: bokeh graph showing the rusult of models being opened with "audit".
    if configured an email will be sent if the model is corrupt. see readme for email config guide in notify/email/
    success: green, unclassified error:orange, corrupt model: red.
    run separately from process_model with: "python" from commands/pulse/ directory.


how to extend:

  • rvt_model_services was created with extendibility in mind. if you wish to extend the functionality, just copy cookiecutter command and adjust or override.


  • of journal files:
    • no white spaces in model path
    • no non-ascii characters in model path
    • task will not run to completion if confronted with any unexpected messages
  • of this framework:
    • do not schedule overlapping readouts, the simple pandas code is not (yet) prepared to handle this in the graphs
    • if you want to use windows network paths use the --nofilecheck option


note: If you are not on this list, but believe you should be, please contact me!



You can’t perform that action at this time.