netcore

[dosymep]

No description provided.

[dosymep]

@eirannejad, I don't have libs on .NET Core

• FileFormatWavefront.dll
• Ifc.NET.dll
• Ifc.NET.XmlSerializers.dll
• Microsoft.WindowsAPICodePack.dll
• Microsoft.WindowsAPICodePack.Shell.dll
• pyRevitLabs.MahAppsMetro.dll (available alpha version)
• PythonStubsBuilder.dll
• Rhino3dmIO.dll
• rhino3dmio_native.dll
• MadMilkman.Ini
• Expression.Blend.Sdk

[dosymep]

• transferred pyRevit paths to targets
• added revit 2025 dependensis
• added revit 2026-2025 docs (xml) (removed)
• added revit 2025 project to pyRevit.Runtime
• added revits preprocessor variables like this
• added multitargets net48, net8.0
• remove $(RevitVersion) from version pyRevitLoader

[dosymep]

I think we need to divide projects into those that use Revit, just libraries, applications.
Transfer all libraries to nuget, thereby greatly simplifying the assembly of the solution.
Remove destination paths from the solution and compile using dotnet build -c Release -o "pyRevit/bin".

[RTLflat]

One thing I noticed when I attach the visual studio debugger: I got a ton of Exception thrown: 'IronPython.Runtime.UnboundNameException' in pyRevitLabs.IronPython.dll messages, and the loading takes ages.

@sanzoghenzo
This is your Visual Studio debug settings. Does this as well in pyRevit 4. Can't rembember which debug setting causes it, but I can show what my current debug settings are that don't have the issue if you want

[sanzoghenzo]

Alright, Cpython tests are working now! I was overwriting the __builtins__ of the module with the pyrevit variables instead of updating the dict 😅

I also added a config parameter for the pyrevit build products command that gets passed to dotnet build, so that one can build everything (well, except the dependencies) in Debug mode (pyrevit build products debug) and attach the visual studio debugger.

The last commit is a refactor of the various Get.*Engine.* methods to use linq and simplify the netcore/netfx conditions; maybe it will solve @jmcouffin problem, but my expectations are very low...

[sanzoghenzo]

@jmcouffin you were right, I've edited too many lines of code and introduced a bug 😅

In trying to reduce the code repetition, I wrongly set all the engines as netfx, even the netcore ones. It is now fixed.

The code in pyrevit cli throws an exception, but the program itself exits with a 0 exit code, that's why the installer doesn't complain.
I think this should be changed, and use an non-zero exit code if there's any error. this also helps when using scripts to automate things (I'm thinking about corporate setup and configuration)

EDIT: I went ahead and added the non-zero exit code. That's it for today, I consider it mission accomplished 😎

[jmcouffin]

mission accomplished

Clearly

[jmcouffin]

Checked this morning the latest WIP installer
It works in all versions of Revit!
Good Job @sanzoghenzo and @dosymep 🍾 🚀

---

A thing I noticed, and I am pretty sure this is not a new bug:

==> Running Revit Instances
Debug: Getting host product info for: 20240319_1700(x64)
Debug: Getting data source "C:\Users\Jean-Marc\AppData\Roaming\pyRevit\Cache\pyrevit-hosts.json"
Debug: Getting data source "C:\Users\Jean-Marc\AppData\Roaming\pyRevit-Master\bin\pyrevit-hosts.json"
Debug: Data source exists "C:\Users\Jean-Marc\AppData\Roaming\pyRevit-Master\bin\pyrevit-hosts.json"
Debug: Using already loaded data. Identical signatures "1205207726" = "1205207726"
Error: Object reference not set to an instance of an object. (System.NullReferenceException)
   at pyRevitCLI.PyRevitCLIRevitCmds.<>c.<PrintLocalRevits>b__1_0(RevitProcess x) in D:\a\pyRevit\pyRevit\dev\pyRevitLabs\pyRevitClI\PyRevitCLIRevitCmds.cs:line 31
   at System.Linq.EnumerableSorter`2.ComputeKeys(TElement[] elements, Int32 count)
   at System.Linq.EnumerableSorter`1.ComputeMap(TElement[] elements, Int32 count)
   at System.Linq.EnumerableSorter`1.Sort(TElement[] elements, Int32 count)
   at System.Linq.OrderedEnumerable`1.GetEnumerator()+MoveNext()
   at pyRevitCLI.PyRevitCLIRevitCmds.PrintLocalRevits(Boolean running) in D:\a\pyRevit\pyRevit\dev\pyRevitLabs\pyRevitClI\PyRevitCLIRevitCmds.cs:line 31
   at pyRevitCLI.PyRevitCLIAppCmds.MakeEnvReport(Boolean json) in D:\a\pyRevit\pyRevit\dev\pyRevitLabs\pyRevitClI\PyRevitCLIAppCmds.cs:line 144
   at pyRevitCLI.PyRevitCLI.ProcessArguments() in D:\a\pyRevit\pyRevit\dev\pyRevitLabs\pyRevitClI\PyRevitCLI.cs:line 189
   at pyRevitCLI.PyRevitCLI.Main(String[] args) in D:\a\pyRevit\pyRevit\dev\pyRevitLabs\pyRevitClI\PyRevitCLI.cs:line 150

the hosts file is located in all 4 locations:

which is a bit of a waste considering the CLI is looking for all these location apparently

and checked against in the CLI in the running instances and throws a NullReferenceException
Previously, the behaviour of the CLI was:

• NullReferenceException when no identical version of Revit running was found in the pyrevit-hosts.json file
• Listing properly running instances of Revit when the list pyrevit-hosts.json was up to date and running instances versions where found in the file

---

Like @dosymep suggested yesterday, I think we could merge this PR and making it a beta release and from now on, address the remaining issues in a new PR.
This might involve adding the GH workflow for publication from your branch @dosymep ?

Thoughts?

[sanzoghenzo]

Yes, I think this is ready to be merged.
If you don't mind, I would rebase and squash all the commits into a single one, so that the git history remains clean.

[dosymep]

We need to merge the current master into my branch, and then make a squash commit. Then we'll refine the beta and fix the problems in the new PR. I don't have the opportunity to do this at the moment, I'm far from home :)

good work @sanzoghenzo @jmcouffin :)

[jmcouffin]

We need to merge the current master into my branch, and then make a squash commit. Then we'll refine the beta and fix the problems in the new PR. I don't have the opportunity to do this at the moment, I'm far from home :)

good work @sanzoghenzo @jmcouffin :)

You do it, or I do? @sanzoghenzo

[sanzoghenzo]

You do it, or I do? @sanzoghenzo

My OCD will make me take way longer than needed, so you can go ahead, even without squash (the git history is a total mess anyway 🤣)

Only one thing: maybe we should merge the develop-4 branch onto dosymep's, so that we have also the latest fixes...

[jmcouffin]

@dosymep I will need write access to your fork to make this happen
dosymep#2

[jmcouffin]

Only one thing: maybe we should merge the develop-4 branch onto dosymep's, so that we have also the latest fixes...

exactly what I thought

[sanzoghenzo]

@dosymep I will need write access to your fork to make this happen dosymep#2

You could have done it directly in your local git without the need for PRs, just add dosymeps repo as a remote im your working folder

[jmcouffin]

I did my best to squash and merge it without messing the work previously done 🤞
I will merge this one so that we start clean with bug fixing and new items

[sanzoghenzo]

The merge of develop-4 deleted the .cs files in the dev/pyRevitLabs.PyRevitRuntime directory!

[jmcouffin]

The merge of develop-4 deleted the .cs files in the dev/pyRevitLabs.PyRevitRuntime directory!

It might be a mistake from my side, let me check

[jmcouffin]

The merge of develop-4 deleted the .cs files in the dev/pyRevitLabs.PyRevitRuntime directory!

do we need those? with all the changes?

[sanzoghenzo]

Yes, the .cs files are needed, otherwise the resulting pyRevitLabs.Pyrevit.Runtime.YYYY.dll would be empty!

The build and CI doesn't fail because the Directory.Build.props in the parent directory has <Compile Include="..\*.cs" />, so it doesn't look for specific files to be present

[jmcouffin]

Yes, the .cs files are needed, otherwise the resulting pyRevitLabs.Pyrevit.Runtime.YYYY.dll would be empty!

The build and CI doesn't fail because the Directory.Build.props in the parent directory has <Compile Include="..\*.cs" />, so it doesn't look for specific files to be present

fixing it right now

[jmcouffin]

@sanzoghenzo can you go on slack, that will make things easier?

[eirannejad]

Y'all are rockstars!