HOOPS Visualize can load a variety of different 2D and 3D file formats. If HOOPS Exchange is also licensed, the list of importable file formats is extended even further. Regardless of which type of file you are loading, HOOPS Visualize uses a common pattern to load all files. There are four basic steps:
- Use an import options kit to set the file load options.
- Load the file. Using a notifier is highly recommended, but not required. All files are loaded asynchronously.
- Take appropriate action based on the status of the notifier.
- Respond to any errors or exceptions.
General import example
The following example loads a HOOPS Stream File (HSF). HSF is the native file format for Visualize. An HSF contains all model information as well as its related scene hierarchy that can be used to save and load complete scenes. The data is stored in a compressed format. HSFs are loaded using the Stream class. Other file formats use a similar class. For example, to load an STL file, you would use the STL class. It is recommended to make use of a notifier during file loading. Notifiers are available for all file types - their purpose is to track loading progress:
This snippet demonstrates the use of Wait() on your notifier object. Because all notifier classes do their I/O in a separate thread, the potential exists to use the object before it is done loading, especially if you try to interact with it immediately after you call Import(). Calling Wait() ensures the I/O operation is complete before continuing. If you do not use Wait(), you could end up with unexpected behavior.
After the file load is finished, you should test the result of the load operation to ensure there was not an unexpected failure:
Assuming the file was loaded correctly, you can get information about the model by querying the associated HPS::Stream::ImportResultsKit.
NOTE: For HSF files, any user options contained within will be UTF8 encoded and stored as user data.
IMPORTANT: During the import process, if you did not use the HPS::Stream::ImportOptionsKit to specify a destination segment, Visualize will create a root segment and place the model there. In this case, to display the model it is necessary to call ShowSegment (as shown in the previous code snippet) in order to get the key of the segment where the model was loaded. After locating the model, it can be included into the tree for rendering. See HPS::Stream::ImportOptionsKit::SetAlternateRoot.
This general behavior also applies to portfolios. If a portfolio is not specified using HPS::Stream::ImportOptionsKit::SetPortfolio, Visualize will create a portfolio for any definitions that need to be created during the import. The location of the portfolio can be shown by calling HPS::Stream::ImportResultsKit::ShowPortfolio.
If the file import is taking too much time, the user changes his mind about loading the file, or your application detects it is running out of memory, you have the ability to cancel a file import.
Canceling is not immediate. Once you cancel, a notice is posted that the import needs to be cancelled. The importer checks periodically to see if the notice has been posted, and when it sees it, it cancels itself.
Importing other file formats
In addition to HSF files, HOOPS Visualize can also natively import STL, OBJ, and point cloud files. The procedure follows the same pattern as HSF. As before, be sure to set the destination segment in the importer's options kit:
Notes on point cloud files
The point cloud importer follows the same pattern. The only real option of note is HPS::PointCloud::ImportOptionsKit::SetPointColor. One thing to be aware of is if the point cloud file specifies RGB colors for the points, the point color value you specify is ignored. If the file specifies no color, the color from SetPointColor will be set as the vertex color in the containing segment. If the file specifies a grayscale intensity for the point colors, each resolved color will be the supplied color multiplied by the intensity for the given point. If the file specifies colors or intensities then the colors will be inserted as vertex parameters. If the file only contains points, then the color will be set as an attribute on the containing segment.
Point clouds are inserted as vertex-only shells (in batches of 10000 points per shell).
Notes on MTL
For MTL files, the OBJ importer opens and reads them in order to discover which material properties are present in the file. The file is closed when the import completes or is interrupted by an exception.
HOOPS Exchange and HOOPS Publish
Importing 2D images
The examples from previous sections have all dealt with loading 3D CAD files. 2D images are loaded slightly differently. First of all, the process uses the HPS::Image namespace. Secondly, the result of the file load is an HPS::ImageKit. Normally the image is loaded from an external file, although it is possible to use a programmtically-defined or generated image as well. This is the process you would use to load an image to prepare it for use, for example, as a texture:
Visualize can load common image formats, including JPG, PNG, and TGA. The full list of supported file formats is provided in the appendix. It is also possible to save 2D images out of an HPS::ImageKit in a similar way. Note that the size and export format must be specified within the HPS::ImageKit itself.
In order to save a screenshot, you can skip the HPS::ImageKit entirely and pass a HPS::WindowKey as a parameter. If you do not specify the image size, the window size is used. If you do specify the image size, and it does not match the window size, the resulting image will be stretched to match the size you specify. This procedure is as follows:
Import events are triggered by Visualize as part of the HSF file loading process. Import events are useful when you need to perform additional processing for individual geometry and attributes. Each event is triggered before Visualize inserts the object into the database so that your logic can decide how to proceed. Each type of import object is represented by a particular HPS::Stream::ImportEvent. For example, if Visualize reads a HPS::CylinderKey from an HSF file, it will trigger a HPS::Stream::CylinderImportEvent. If changes are about to be made to visibility attributes during the HSF file read, a HPS::Stream::VisibilityImportEvent will be generated.
You receive these events by setting an HPS::Stream::ImportEventHandler for the HPS::Stream::ImportEvent type on your HPS::Stream::ImportOptionsKit,. The HPS::Stream::ImportEventHandler class has a virtual function, HPS::Stream::ImportEventHandler::Handle, that Visualize will call with the HPS::Stream::ImportEvent data. You should override the Handle function to inspect or modify the data to suit your needs. Each event is triggered before the database is updated, and your implementation of Handle will have the ability to determine whether the import operation should proceed. If Handle returns true, the operation will proceed, false means the operation will be discarded.
The following code sample defines a custom HPS::Stream::ImportEventHandler class, which will unset any text 'bold' attributes that might be present in the HSF file:
We then set the custom handler on the HPS::Stream::ImportOptionsKit for the two different text import events, prior to importing the HSF File:
There are two different ways of storing user data in an HSF file. One form is user data that is stored within the database and is associated with a segment or geometry (See user data), and will be reported to you via an HPS::Stream::UserDataImportEvent. The second form is user data that is not associated with the database and will be reported to you via an HPS::Stream::NonDBUserDataImportEvent. (Note: HOOPS Visualize does not currently provide support for exporting non-DB user data, but existing HSF files may contain such user-data.