| 11:56.42 | *** join/#stepcode infobot (ibot@rikers.org) |
| 11:56.42 | *** topic/#stepcode is STEPcode - data exchange with the technologies of ISO10303 || http://stepcode.org || LOG http://ibot.rikers.org/%23stepcode/ |
| 13:02.34 | *** join/#stepcode metal-machine (~metal-mac@117.205.75.45) |
| 13:03.25 | metal-machine | Hi, I am trying hard to know this, Can anyone help me how to play with IFC files using Stepcode? |
| 13:03.40 | metal-machine | Any example code or something? |
| 14:59.11 | *** join/#stepcode metal-machine (~metal-mac@117.205.75.45) |
| 15:03.16 | metal-machine | ping |
| 15:17.04 | starseeker | metal-machine: Does IFC have a step EXPRESS schema? |
| 15:17.57 | metal-machine | these are based on same. no? |
| 15:18.48 | starseeker | Well, the normal way you work with stepcode is you feed an express schema to exp2cxx, and then compile the generated code and the libraries in stepcode together with your import code to make an importer |
| 15:19.18 | starseeker | metal-machine: are you familiar with this work on IFC? http://yorik.uncreated.net/guestblog.php?tag=freecad |
| 15:19.28 | metal-machine | yes |
| 15:19.49 | metal-machine | i am playing with it from almost last 3 weeks |
| 15:19.51 | starseeker | there is also ifcOpenShell |
| 15:20.18 | metal-machine | and had conversation with ifcOpenshell developer |
| 15:20.41 | metal-machine | I told them that I want to IFC to .g converter |
| 15:20.49 | starseeker | Ah. |
| 15:21.23 | starseeker | OK, the best I can suggest is to look at our step-g importer in src/conv/step/step-g in BRL-CAD |
| 15:21.49 | metal-machine | yes i have also compiled step-g converter on my pc |
| 15:21.55 | metal-machine | and step code also :) |
| 15:22.35 | metal-machine | looked the stepcode docs but unable to find it's usage with IFC cooking |
| 15:22.39 | starseeker | OK. So the next step then would be to make an ifc-g directory, and duplicate the step-g CMake logic |
| 15:22.52 | starseeker | then you would replace the step AP203 schema with the IFC schema |
| 15:23.01 | starseeker | is hunting for the IFC schema now... |
| 15:23.57 | metal-machine | sorry I am unable to understand that last hunting line |
| 15:24.28 | starseeker | ah, stepcode has it |
| 15:24.40 | starseeker | stepcode/data/ifc4 |
| 15:24.44 | starseeker | was looking for that |
| 15:24.52 | metal-machine | yes while compiling I saw stepcode has ifc4 |
| 15:24.57 | metal-machine | latest |
| 15:25.02 | metal-machine | IFC schema |
| 15:25.03 | starseeker | OK. |
| 15:25.49 | starseeker | so you would use that file instead of data/ap203/ap203.exp |
| 15:26.14 | metal-machine | in which part of code? |
| 15:26.27 | starseeker | metal-machine: in the ifc-g directory you would create |
| 15:26.51 | starseeker | metal-machine: if it would be simpler, I could take a few minutes and set up some basic machinery in BRL-CAD |
| 15:26.52 | metal-machine | I mean can you tell me the flow of code that how it works |
| 15:27.30 | starseeker | Oh. Well, the IFC4.exp file will generate a set of headers and cpp files, which express the data types defined in the schema |
| 15:27.54 | metal-machine | yeah ifcopenshell also work like that |
| 15:28.53 | starseeker | our STEPWrapper header/cpp file are convenience wrappings around the stepcode functionality for loading files |
| 15:29.15 | starseeker | so if you look at step-g.cpp, you will see the STEPWrapper being used to load (in our case) a step file |
| 15:29.30 | metal-machine | ok |
| 15:29.53 | metal-machine | ok so in my case I have to load IFC file |
| 15:29.57 | starseeker | you would need to do something similar to load an IFC file (wouldn't necessarily have to do it our way with the wrapper, but similar steps would be needed) |
| 15:30.00 | starseeker | exactly |
| 15:30.35 | metal-machine | yeah that seems interesting |
| 15:30.38 | starseeker | once loaded, you would need to see how to "unpack" the IFC data parsed into the C++ structures generated by stepcode and populated by the import |
| 15:31.08 | starseeker | once you had access to that data, you would then need to define and implement a translation from their data into BRL-CAD's geometry |
| 15:31.25 | starseeker | (again, step-g creates a .g file so the basics are there) |
| 15:31.37 | metal-machine | ok so MGED will understand that? |
| 15:31.44 | starseeker | the .g file? yes |
| 15:32.07 | metal-machine | yes.. |
| 15:32.32 | metal-machine | one more thing about .g file. it is king of binary format for BRL-CAD? |
| 15:32.37 | starseeker | once the data is translated from IFC concepts to BRL-CAD concepts and written into a .g file, it is ready for opeing by BRL-CAD |
| 15:32.43 | starseeker | yes, binary format |
| 15:32.53 | starseeker | we have an API that you would use to create entities in it |
| 15:32.59 | starseeker | librt and/or libwdb |
| 15:33.20 | starseeker | for simpler examples of creating .g files, take a look at src/shapes |
| 15:33.25 | metal-machine | ah :) wow that is lot of help. I mean amazing! |
| 15:34.21 | starseeker | metal-machine: all the hard work is in figuring out how to access the IFC data as parsed by stepcode and identifying the mappings between their data and the BRL-CAD entities |
| 15:35.00 | starseeker | if IFC is anything like the AP2** STEP standards, you'll have a bit of a maze to follow |
| 15:35.10 | metal-machine | yesI guess IFCOpenshell code reading can help in that |
| 15:35.27 | starseeker | possibly |
| 15:36.05 | starseeker | you'll have to really understand how information is expressed in IFC - FreeCAD's work may actually help a bit there |
| 15:36.42 | metal-machine | yes I am reading the IFC data model things from some days |
| 15:37.07 | metal-machine | and HS Rai is my mentor :) |
| 15:37.17 | starseeker | very good :-) |
| 15:38.11 | starseeker | I would suggest stubbing in a bare-bones ifc-g directory in src/conv and studying how the step-g directory is set up |
| 15:38.35 | starseeker | start by making a very simple program to just open an ifc file and print out all the data in it |
| 15:38.47 | metal-machine | ok time to play with things. :) |
| 15:38.49 | starseeker | (i.e. don't worry about .g translation until you can access the ifc data) |
| 15:38.58 | metal-machine | yeah |
| 15:39.49 | starseeker | some of the macros defined in src/conv/step/CMakeLists.txt will probably be useful |
| 15:40.25 | metal-machine | anything perticular stuff about IFC you feel I should go(just for better understanding of data model) |
| 15:40.44 | starseeker | can't help much there - I don't know a lot about IFC |
| 15:40.51 | metal-machine | okay |
| 15:41.05 | starseeker | if you like I can set up a basic ifc-g directory to save you that grief |
| 15:42.01 | metal-machine | :) |
| 15:42.32 | starseeker | give me a few minutes |
| 15:43.32 | metal-machine | yeah sure, in mean time exploring the things you told |
| 16:17.20 | starseeker | metal-machine: done - see r63581 |
| 16:19.29 | starseeker | metal-machine: what's there will do a basic load of an ifc file and report what's there - I tested it out using the testhouse.ifc file from the FreeCAD work |
| 16:19.54 | starseeker | you'll need to add the -DBRLCAD_ENABLE_IFC=ON option to CMake to build |
| 16:20.15 | metal-machine | ok yes I am looking |