Thank you so much! I couldn't ask for a more thorough response.

Dan

@ferdinand cool. this really helpful!

Hey @moghurt,

Thank you for reaching out to us.

I tried using the pyp plugin to execute Python scripts when starting the software to modify the project file, but when I use the '-render' parameter, the Python code in the pyp plugin is executed after the rendering is complete.

I can only speculate as to what exactly you have been doing here, as you do not provide any details or code. Please have a look at Support Procedures: Asking Questions for how to ask good technical questions.

The general workflow for what you want to do here (given that one really wants to follow the plugin route) would be to implement PluginMessage for your plugin, and then try to run your code from there at a fitting point of time in the startup sequence of Cinema 4D. And while I would not rule out that -render has an impact on the startup sequence of Cinema 4D, the general claim that your Python code only runs after the rendering seems a bit far fetched. Because for events such as C4DPL_INIT_SYS and C4DPL_INIT modules and plugins have not been loaded yet, i.e., rendering is literally impossible at these points.

The plugin approach also assumes that you can inject without problems alien command line arguments into a Cinema 4D app call, e.g., \CINEMA 4D.exe"-nogui -render "render.c4d" -script "myscript.py" with -script being the alien argument. I am not quite sure that this would work, I would have to try myself. What also is problematic, is that you are likely not able to get hold of the render document on this route. The active document will be the empty document, not the render document, i.e., you cannot get hold of the BaseDocument for render.c4d. And you cannot just open and modify the document yourself, as Cinema might have already a lock on that file or is even already loading it.

All this makes using c4dpy instead much more attractive as you won't have these problems there. It is also the intended solution for what you want to do here: Modify one or many documents without a GUI in a fire and forget manner. A while ago we had the topic Extending the Command Line with Python where I showed a workflow for how to treat c4dpy as a Script Manager environment.

ℹ The thread is four years old, written in Python 2 syntax (you will have to convert to Python 3 syntax), and I was then not yet a Maxon employee but a student. So, you might run into some issues. But the principle is definitively reproduceable and I also do not see anything in my old code which seems like a really bad idea.

Your workflow could then look like this for doing the pre and post rendering events (as a bat, ps1, sh, or command file and assuming that you do not want to overwrite files):

c4dpy script_manager_environment.py -script my_script.py -in my_doc.c4d -out my_doc_pre.c4d CommandLine -my_doc_pre.c4d ... c4dpy script_manager_environment.py -script my_script.py -in my_doc_pre.c4d -out my_doc_post.c4d

Cheers,
Ferdinand

Yes, I know about not using multiple events. I was playing with that, I thought that's the reason the button was not being called

Thank you @ferdinand

hey,

found the error. works exactly as i had planned. my code was correct, but i forgot to switch off use fields on the new vertex i mapped.

thanks for the help anyway!

grafik.png

Thank you, @ferdinand. I modified sourceprocessor.py for now, and it works.

Hey @bentraje,

I am glad that you found your solution, and yes, you are absolutely right, it should be GeGetCurrentThread . I fixed that in my pseudo-code.

Cheers,
Ferdiand

@ferdinand

yea python is much more forgiving than C++. i was just wondering if there was an already existing API for such task. 👍

Good to hear!

@i_mazlov , Do you speak Russian?

Hey @GordOrb,

This is not possible. You are not really meant to distinguish the file names you are writing for. I also do not understand the purpose of doing what you want to do. You can already name the beauty file and the multipass file differently in the UI. And when you write out each pass on its own, Cinema 4D will already name them differently for you.

There will be singular token call for all pass layers, and before that a call for the RGB/beauty layer. None of the calls for naming the file contains pass information. The inner logic of it seems to be that you first get the call for evaluating the token for a filename and after that the actual passes.

The purpose of some fields in the passed data is probably more of internal than public nature, since you cannot really do much with that info due to how the token calls are coming in. In some contexts these fields might be useful when you use the Token System (C++ Manual) more manually. But at least for the passes case, this is not really true either, because when you start to render manually, you will have to fill in the pass data on your own when constructing a RenderPathData.

So, long story short: this is not really possible, at least in the intended way. Upon having a closer look at the time stamps and my file output, I realized that you could try to rely on the rule 'the call after the call for the RGB pass', because that is at least how it works in my example.

isMultipass: '1', pname: ' ', ptype: ' ', pid: ' -1', t: 9366346949848 // First call, irrelevant isMultipass: '1', pname: ' ', ptype: ' ', pid: ' -1', t: 9367113676312 // Time stamp of saved RGB/beauty. isMultipass: '1', pname: ' RGB', ptype: ' RGB', pid: ' -1', t: 9367113987757 // RGB/beauty pass isMultipass: '1', pname: ' ', ptype: ' ', pid: ' -1', t: 9367114252401 // Time stamp of all passes. ...

If this works, this would be hack, and it would be then up to you establish if that holds true or not. To establish "what is the last call", you could have some global list into which you push things. Make sure not to store any Cinema 4D rendering data in that global list (BaseDocument, BaseContainer, RenderData) or you might end up in world of hurt.

Cheers,
Ferdinand

Written Files:
fc970227-90fb-437e-b5d7-b7a25a9e2da0-image.png
My console dump, here I search for the time stamp 9367114252401 of my files, it is the fourth call, long before we get the calls for the passes. The only thing that comes before is a call for the RGB, i. e., the beauty pass.
b5e92449-2ffa-4d4b-a0b7-de02968635d8-image.png
Code:

""" """ import c4d import mxutils import time def get_render_type(data: c4d.BaseContainer) -> str: """Dumps information about #data into the token and the console. """ # Get the render document #rdoc, and build a string for the UUID of rdoc as well as its memory # location and #data. Unsurprisingly, the mem locs are meaningless here in Python as everything # must be wrapped by Python objects in each call. In C++ you might be able to establish "sameness" # between calls by the location of #data (the rendering document which is used is always the # same as shown by its UUID). But I doubt that you will have luck with #data in C++, because it # is likely constructed for each call there too (and not reused). rdoc: c4d.documents.BaseDocument = mxutils.CheckType(data[0], c4d.documents.BaseDocument) uuid: bytes = bytes(rdoc.FindUniqueID(c4d.MAXON_CREATOR_ID)) ident: str = f"doc-uuid: {uuid}, doc-mem: {id(rdoc)}, bc-mem: {id(data)}" # Build the token. token: str = (f"isMultipass: '{rdoc.GetActiveRenderData()[c4d.RDATA_MULTIPASS_ENABLE]}', " f"pname: '{data[5]:>20}', ptype: '{data[6]:>10}', pid: '{data[7]:>3}', t: {time.perf_counter_ns()}") print (ident, token) return token if __name__ == "__main__": if "r_type" not in set(item["_token"] for item in c4d.modules.tokensystem.GetAllTokenEntries()): c4d.plugins.RegisterToken("r_type", "[Render] Type", "reg/multi", get_render_type)

Hey @Visualride-0,

you seem to run into some form of memory leak. That you open that many files in a row on one Cinema 4D instance is of course not really intended. This could either be a l smaller leak in our API that simply manifests more or only at all when you load that many files.

Generally, you could separate out the file iteration into an external script, a shell script or a vanilla Python script if you want, and then simply run c4dpy with your script per file or for batches of 10 or 100 files. Starting and shutting down c4dpy 8000 times in a row will of course take its toll in time, so one would be incentivized to bundle up more files to bring down that number of starts and shutdowns.

I had a look at your code, and apart from your redundant document insertion, I do not see much wrong with it, as it is basically just the loop on fileList. Cinema 4D should have caught your mistake there and prevented you from opening the same file twice, but I have not tested it.

for f in filesList: # This line will load #f into a BaseDocument, insert it into the opened files, and make it the # active document. Unlike for BaseDocument.LoadDocument, the following lines are therefore # redundant. c4d.documents.LoadFile(f) # doc = c4d.documents.GetActiveDocument() # c4d.documents.InsertBaseDocument(doc) # c4d.documents.SetActiveDocument(doc) c4d.documents.SaveDocument(c4d.documents.GetActiveDocument(), f, c4d.SAVEDOCUMENTFLAGS_0, c4d.FORMAT_C4DEXPORT) # This does what it should do, close the active document, and create a new one. But to be # ultra safe, you could also call c4d.documents.CloseAllDocuments(). c4d.documents.KillDocument(doc)

What escapes mea bit too, is the purpose of what we are doing here, since we just overwrite files with itself. But I guess this is minimized code, so it is only meant to demonstrate the problem. For my own understanding, the files in fileList are all *.c4d files, right?

Fixing your code as shown above will likely not fix the problem, it seems more likely to me that there is a memory leak somewhere. For an intermediate solution, you will probably have to do what I proposed above, split this into batches of Cinema 4D or c4dpy instances to flush your memory by shutting down Cinema 4D.

But we would like to have a look at the problem. Could you provide a sample of the documents, something between 10 or 100 files, so that we can test it with your data? You can send me a cloud storage link via a PM on the forum or via a mail to sdk_support(at)maxon(dot)net. Mail attachments will not work here due to the 50MB limitation.

We strive for Cinema 4D being memory leak free, and if there is a severe leak we are not aware of, we should fix it. But this might take time.

Cheers,
Ferdinand

Hi Maxime, thank you very much for your answer! I wasn't sure if storing things in the global document BaseContainer was a good approach, but it's working well.

Cheers,
Daniel

Hi,
thank you for the tips.

Here is my C++ -treegenerator-plugin, written for R25:

http://www.klausfilm.bplaced.net/TreeGenerator/Website_Treegenerator.html

Cheers
Klaus

Thanks for all the tips @ferdinand

The solution was Catalina + XCode 12.4

Hi @ferdinand,

Thank you so much for the comprehensive reply! Given all this I don't think it would make sense to keep trying to unit test this particular part of our code with Python. The last approach you highlighted with testing directly on the C++ side seems like the best option, so I'll give that a try. I'll have to see how to integrate this test with the rest of the unit tests (we have a lot and they're all in Python) but hopefully that won't be too much of a problem.

Thanks again for all the help,
Daniel

Hey @Madara,

in short, a Python generator object is not meant to generate splines. You can collapse the cache of your constructed spline and return its LineObject cache instead of the spline itself. For an example, see [1]. In your case that would be connect_obj. This will work in the sense that the generator cache will be then like a SplineObject cache and when you 'Current State to Object' your generator, it will behave like a spline and give you the interpolated editable linear spline and not a spline generator.

However, spline snapping not only inspects caches, but also checks if OBJECT_ISSPLINE is true. So, you cannot bamboozle it by faking a SplineObject cache. You will have to implement a proper spline plugin for what you want to do, as you can overwrite there GetContour

Cheers,
Ferdinand

PS: Your code contains multiple elements that can crash Cinema 4D and/or corrupt the loaded document. Specifically, these sections:

if link_clone: #Parametric object pobj = u.SendModelingCommand( command = c4d.MCOMMAND_CURRENTSTATETOOBJECT, list = [connect_obj], mode = c4d.MODELINGCOMMANDMODE_ALL, doc = op.GetMain()) connect_obj = pobj[0] ... offspline = u.SendModelingCommand( c4d.MCOMMAND_SPLINE_CREATEOUTLINE, [connect_obj], c4d.MODELINGCOMMANDMODE_ALL, bc=settings, doc=doc)

A Python generator's main() function is subject to the threading restrictions of Cinema 4D because it is just a wrapper for ObjectData.GetVirtualObjects. Executing SendModelingCommand in a threaded context is fine, but you cannot do it on the active document (both doc and op.GetDocument() are the active document). This is here even worsened by the fact that that the objects in list are not actually part of the document you pass as doc.

To circumvent that you must create your own document just as I have in [1] with temp. Find a more detailed example of that technique in smc_extrude_s26.py.

[1]

import c4d op: c4d.BaseObject # The Python generator object. def main() -> c4d.BaseObject: """ """ # A user data Boolean to toggle collapsing the cache of this generator. collapse: bool = op[c4d.ID_USERDATA, 1] spline: c4d.BaseObject = c4d.BaseObject(c4d.Osplinecircle) if not spline: return c4d.BaseObject(c4d.Onull) # Just return the circle spline itself when #collapse is false. if not collapse: return spline # Otherwise grab the LineObject cache of the spline and return a copy of it. temp: c4d.documents.BaseDocument = c4d.documents.BaseDocument() temp.InsertObject(spline) temp.ExecutePasses(c4d.threading.GeGetCurrentThread(), False, False, True, c4d.BUILDFLAGS_NONE) cache: c4d.LineObject | None = spline.GetCache() if not isinstance(cache, c4d.LineObject): return c4d.BaseObject(c4d.Onull) return cache.GetClone(c4d.COPYFLAGS_NONE)

super, thanks

@jana sorry, I forgot to flag it myself. All solved, thank you 👍

Hello @Ryan_Territory,

Welcome to the Plugin Café forum and the Cinema 4D development community, it is great to have you with us!

Getting Started

Before creating your next postings, we would recommend making yourself accustomed with our Forum and Support Guidelines, as they line out details about the Maxon SDK Group support procedures. Of special importance are:

Support Procedures: Scope of Support: Lines out the things we will do and what we will not do. Support Procedures: Confidential Data: Most questions should be accompanied by code but code cannot always be shared publicly. This section explains how to share code confidentially with Maxon. Forum Structure and Features: Lines out how the forum works. Structure of a Question: Lines out how to ask a good technical question. It is not mandatory to follow this exactly, but you should follow the idea of keeping things short and mentioning your primary question in a clear manner. About your First Question

This exact question of yours has been asked before here for C++. Find a Python variant of the code below. For the technical details, I would recommend reading the older thread.

Cheers,
Ferdinand

Result:
557e9ddd-793a-4d07-aa6e-5ffd7ad05687-image.png

Code:

import c4d # Define #Orslight, only must be done in pre 2023.0 API versions. c4d.Orslight: int = 1036751 def main(): doc = c4d.documents.GetActiveDocument() # Create Light light = c4d.BaseObject(c4d.Orslight) # Set the attributes light[c4d.REDSHIFT_LIGHT_TYPE] = c4d.REDSHIFT_LIGHT_TYPE_DOME # Define the texture parameter ID. texturePathId: c4d.DescID = c4d.DescID( # 1st DescLevel for the outer unexposed datatype of "Texture". c4d.DescLevel( c4d.REDSHIFT_LIGHT_DOME_TEX0, # The parameter ID (12000) for "Texture" 1036765, # The datatype ID for for "Texture", this type is not exposed, # we must use the raw ID. c4d.Orslight # The creator ID, Orslight in this case. ), # 2nd DescLevel for the inner exposed datatype of "Texture.Path". c4d.DescLevel( c4d.REDSHIFT_FILE_PATH, # The sub-channel parameter ID (1000) for "Texture.Path" c4d.DTYPE_STRING, # The data type ID of this component, string in this case. 0 # The creator ID, I just went with undefined here. )) light[texturePathId] = "asset:///file_c622e9ad3ea0e159" # Insert the light into the scene doc.InsertObject(light) c4d.EventAdd() if __name__=='__main__': main()