README_USERD_1.0 ================ -------------------------------------- EnSight User Defined Reader Capability ===> (API 1.0) -------------------------------------- A user defined reader capability is included in EnSight which can allow otherwise unsupported structured or unstructured data to be read. The user defined reader capability utilizes dynamic shared libraries composed of routines defined in this document but produced by you, the user, (or some third party). This capability is currently available for dec, ibm, hp, sgi, sun, linux, alpha linux, and NT servers. Two versions of this API are available starting with EnSight Version 7.2. The 1.0 API (which was designed to be friendly to those producing it, but requires more manipulation internally by EnSight) may be a little easier to produce, but requires more memory and processing time. The 2.0 API is considerably more efficient, and was designed more with that in mind. It requires that all data be provided on a part basis. If you already have a working 1.0 API reader and are happy with it - there is probably no reason to modify it to the 2.0 API unless: - you deal with large models and the memory use and load times are a problem or - you need tensor variable support or - you need complex variable support or - you need multiple timeset capability or _ you want to provide your own "border" elements (as opposed to EnSight's computation of them). If you are producing a new reader, you should consider which will work best for your needs. API 1.0 (defined in this README_USERD_1.0 document) ======= The original user defined reader API (used with EnSight Versions 6 through 7.1) will continue to be supported. (Note that there was a change in the way that the libraries were made at version 7.1 of EnSight, but underlying code was kept the same.) Thus, any readers that work with EnSight 7.1, should still function with EnSight 7.2. API 2.0 (defined in README_USERD_2.0 document) ======= This new API has been defined to be more efficient and includes access to new capabilities of EnSight 7.2. It lends itself closely to the EnSight "gold" type format. Some of its advantages are:: * Most intermediate temporary arrays have been eliminated, such that the user defined routines write directly into internal part structures. This is a considerable improvement in memory use, and improves speed as well since far less memory need be allocated, initialized, etc. * Parts are self contained. Coordinates, connectivity and all variables are provided on a part basis. This eliminates the need for several global to local coordinate mapping operations and the need for node id connectivity hashing. This can greatly improve the speed at which models are loaded. * Model extents can be provided directly, such that EnSight need not read all the coordinate data at load time. * Tensor variables are supported * Complex variables are supported * A routine is provided as EnSight exits, so cleanup operations such as removing temporary files can be easily accomplished. * Geometry and variables can be provided on different time lines. * If your data format already provides boundary shell information, you can use it instead of the "border" representation that EnSight would compute. Further discussion on the philosophical differences between the two API's and an efficiency comparison example can be found in the README_1.0_to_2.0 file. This file also contains guidance on necessary changes to modify an existing 1.0 API to the new 2.0 API. **************************************************************************** Note: Several (1.0 API) user defined readers have been included with your EnSight release and are configured by default. There are site- and user-configurable options outlined in step 3 below. Please be aware that these are "unsupported" readers, but many of them are being used successfully. **************************************************************************** The process for producing a user defined reader is: --------------------------------------------------- 1. Write code for all pertinent routines in the library (Unless someone else has done this for you). This is of course where the work is done by the user. The word "pertinent" is used because depending on the nature of the data, some of the routines in the library may be dummy routines. The source code for a dummy library and for various other working or sample libraries is copied from the installation CD during installation. These will be located in directories under: $ENSIGHT7_HOME/user_defined_src/readers examples of API 1.0: ------------------- Basic dummy routines provide skeleton for a new reader $ENSIGHT7_HOME/user_defined_src/readers/dummy Sample library which reads unstructured binary EnSight6 data $ENSIGHT7_HOME/user_defined_src/readers/ensight6 Sample library which reads binary static plot3d data $ENSIGHT7_HOME/user_defined_src/readers/plot3d Reads binary LS-DYNA3D state database $ENSIGHT7_HOME/user_defined_src/readers/ls-dyna3d Reads FORTRAN binary Unstructured dytran data base $ENSIGHT7_HOME/user_defined_src/readers/dytran Reads FlowScience "flsgrf" flow3d data $ENSIGHT7_HOME/user_defined_src/readers/flow3d Reads Tecplot "plt" files $ENSIGHT7_HOME/user_defined_src/readers/tecplot Reads Common File Format data $ENSIGHT7_HOME/user_defined_src/readers/cff Reads Cobalt grid and picture/restart file data $ENSIGHT7_HOME/user_defined_src/readers/cobalt Reads binary Nastran OP2 data base $ENSIGHT7_HOME/user_defined_src/readers/nastran Reads binary and ascii cfx data $ENSIGHT7_HOME/user_defined_src/readers/cfx4 Reads Exodus II data base $ENSIGHT7_HOME/user_defined_src/readers/exodus Reads Parallel Exodus data base $ENSIGHT7_HOME/user_defined_src/readers/pxi Reads FORTRAN binary SCRYU data $ENSIGHT7_HOME/user_defined_src/readers/scryu Reads binary and ascii STL data $ENSIGHT7_HOME/user_defined_src/readers/stl Reads Vectis data $ENSIGHT7_HOME/user_defined_src/readers/vectis You may find it useful to place your library source in this area as well, but are not limited to this location. * ===> The descriptions of each library routine and the order that the routines are called, which is provided in this file, along with the example libraries, should make it possible for you to produce code for your own data reader. 2. Produce the dynamic shared library. This is a compiling and loading process which varies according to the type of machine you are on. In the user-defined-reader source tree we have tried to isolate the machine dependent parts of the build process using a set of files in the 'config' directory. In this directory there is a configuration file for each platform on which EnSight is supported. Before you can compile the installed readers you should run the script called 'init' in the config directory. i.e. (for UNIX) cd config ./init sgi_6.5_n64 cd .. make If you are compiling for Windows NT, there are two options. If you have the Cygwin GNU utilities installed, you can use GNU make as for Unix. Otherwise, there is a script called makeall.cmd which will build all of the readers using nmake. The Makefiles in each reader directory will work using either make or nmake. i.e. (WIN32 Cygwin) (using nmake) cd config cd config sh init win32 cp win32 config cd .. cd .. mkdir lib make makeall.cmd If you have platform-specific portions of code in your reader, the build system defines a set of flags which can be used within #ifdef ... #endif regions in your source, as shown in the table below. Because the readers are now dynamically opened by EnSight, you may have to include dependent libraries on your link-line to avoid having unresolved symbols. If you are having problems with a reader, start ensight as "ensight7 -readerdbg" and you will get feedback on any problems encountered in loading a reader. If there are unresolved symbols, you need to find the library which contains the missing symbols and link it into your reader by adding it to the example link commands below. If you choose to use a different build environment for your reader, you should take care to use compatible compilation flags to ensure compatibilty with the EnSight executables, most notably on the SGI and HP-UX 11.0 platforms, which should use the following flags: sgi_6.2_o32: -mips2 sgi_6.2_n64: -mips4 -64 sgi_6.5_n32: -mips3 sgi_6.5_n64: -mips4 -64 hp_11.0_32: +DA2.0 hp_11.0_64: +DA2.0W ______________________________________________________________________ | MACHINE | OS flag | SHARED LIBRARY NAME PRODUCED | | TYPE |------------------------------------------------------------| | | LD COMMAND USED IN MAKEFILE | ====================================================================== ______________________________________________________________________ | sgi | -DSGI | libuserd-X.so | | |------------------------------------------------------------| | | ld -shared -all -o libuserd-X.so libuserd-X.o | ---------------------------------------------------------------------- ______________________________________________________________________ | hp | -DHP | libuserd-X.sl | | |------------------------------------------------------------| | | ld -b -o libuserd-X.sl libuserd-X.o | ---------------------------------------------------------------------- ______________________________________________________________________ | sun | -DSUN | libuserd-X.so | | |------------------------------------------------------------| | | ld -G -o libuserd-X.so libuserd-X.o | ---------------------------------------------------------------------- ______________________________________________________________________ | dec | -DDEC | libuserd-X.so | | |------------------------------------------------------------| | | ld -shared -all -o libuserd-X.so libuserd-X.o -lc | ---------------------------------------------------------------------- ______________________________________________________________________ | linux | -DLINUX | libuserd-X.so | | |------------------------------------------------------------| | | ld -shared -o libuserd-X.so libuserd-X.o -lc | ---------------------------------------------------------------------- ______________________________________________________________________ | alpha | -DALINUX | libuserd-X.so | | linux |------------------------------------------------------------| | | ld -shared -o libuserd-X.so libuserd-X.o -lc | ---------------------------------------------------------------------- ______________________________________________________________________ | ibm | -DIBM | libuserd-X.so | | |------------------------------------------------------------| | | ld -G -o libuserd-X.so libuserd-X.o -bnoentry -bexpall -lc | ---------------------------------------------------------------------- Once you have created your library, you should place it in a directory of your choice or in the standard reader location: $ENSIGHT7_HOME/machines/$ENSIGHT7_ARCH/lib_readers For example, if you created a reader for "mydata", you should create the reader libuserd-mydata.so and place the file in your own reader directory (see section 3 below) or in the standard location: $ENSIGHT7_HOME/machines/$ENSIGHT7_ARCH/lib_readers/libuserd-mydata.so 3. By default EnSight will load all readers found in the directory: $ENSIGHT7_HOME/machines/$ENSIGHT7_ARCH/lib_readers Files with names "libuserd-X.so" (where X is a name unique to the reader) are assumed to be user-defined readers. There are two methods which can be used to supplement the default behavior. (1) A feature which is useful for site-level or user-level configuration is the optional environment variable $ENSIGHT7_READER. This variable directs EnSight to load all readers in the specified reader directory (you should probably specify a full path) before loading the built-in readers. If the same reader exists in both directories (as determined by the name returned by USERD_get_name_of_reader(), NOT by the filename), the locally configured reader will take precedence. (2) A useful feature for end-users is the use of the libuserd-devel reader. EnSight will search for a reader named libuserd-devel.so (.sl for HP or .dll for NT). This reader can exist anywhere in the library path (see below) of the user. This is useful for an individual actively developing a reader because the existence of a libuserd-devel library will take precedence over any other library which returns the same name from USERD_get_name_of_reader(). As an example, a site may install commonly used readers in a common location, and users can set the ENSIGHT7_READER variable to access them: setenv ENSIGHT7_READER /usr/local/lib/e7readers A user working on a new reader may compile the reader and place it in a directory specified by the library path: cp libuserd-myreader.so ~/lib/libuserd-devel.so setenv ~/lib:$ The user is responsible for correctly configuring the library path variable in order to make use of the libuserd-devel feature. The library environment variables used are: Machine type Environment variable to set ------------ --------------------------- sgi LD_LIBRARY_PATH dec LD_LIBRARY_PATH sun LD_LIBRARY_PATH linux LD_LIBRARY_PATH alpha linux LD_LIBRARY_PATH hp SHLIB_PATH ibm LIBPATH As always, EnSight support is available if you need it. ------------------------------- Quick Index of Library Routines ------------------------------- Generally Needed for UNSTRUCTURED data -------------------------------------- USERD_get_number_of_global_nodes number of global nodes USERD_get_global_coords global node coordinates USERD_get_global_node_ids global node ids USERD_get_element_connectivities_for_part part's element connectivites USERD_get_element_ids_for_part part's element ids USERD_get_scalar_values global scalar variables USERD_get_vector_values global vector variables Generally Needed for BLOCK data ----------------------------------------- USERD_get_block_coords_by_component block coordinates USERD_get_block_iblanking block iblanking values USERD_get_block_scalar_values block scalar variables USERD_get_block_vector_values_by_component block vector variables Generally needed for either or both kinds of data ------------------------------------------------- USERD_set_filenames filenames entered in GUI USERD_set_time_step current time step USERD_get_name_of_reader name of reader for GUI USERD_get_number_of_files_in_dataset number of files in model USERD_get_dataset_query_file_info info about each model file USERD_get_changing_geometry_status changing geometry? USERD_get_node_label_status node labels? USERD_get_element_label_status element labels? USERD_get_number_of_time_steps number of time steps USERD_get_solution_times solution time values USERD_get_description_lines file associated descrip lines USERD_get_number_of_variables number of variables USERD_get_variable_info variable type/descrip etc. USERD_get_constant_value constant variable's value USERD_get_number_of_model_parts number of model parts USERD_get_part_build_info part type/descrip etc. USERD_get_variable_value_at_specific node's or element's variable value over time USERD_stop_part_building cleanup routine USERD_bkup archive routine ------------------------- Order Routines are called ------------------------- The various main operations are given basically in the order they will be performed. Within each operation, the order the routines will be called is given. 1. Setting name in the gui, and specifying one or two input fields USERD_get_name_of_reader 2. Setting filenames and getting time info USERD_set_filenames USERD_get_number_of_time_steps USERD_get_solution_times USERD_set_time_step 3. Gathering info for part builder USERD_set_time_step USERD_get_changing_geometry_status USERD_get_node_label_status USERD_get_element_label_status USERD_get_number_of_files_in_dataset USERD_get_dataset_query_file_info USERD_get_description_lines (for geometry) USERD_get_number_of_model_parts USERD_get_part_build_info USERD_get_number_global_nodes USERD_get_global_coords (for model extents) USERD_get_block_coords_by_component (for model extents) 4. Gathering Variable info USERD_get_number_of_variables USERD_get_variable_info 5. Part building (per part created) USERD_set_time_step USERD_get_global_coords USERD_get_global_node_ids USERD_get_element_connectivities_for_part USERD_get_element_ids_for_part USERD_get_block_iblanking USERD_get_block_coords_by_component USERD_stop_part_building (only once when part builder dialog is closed) 6. Loading Variables constants: --------- USERD_set_time_step USERD_get_constant_value scalars: ------- USERD_get_description_lines USERD_set_time_step USERD_get_scalar_values USERD_get_block_scalar_values vectors: ------- USERD_get_description_lines USERD_set_time_step USERD_get_vector_values USERD_get_block_vector_values_by_component 7. Changing geometry changing coords only: -------------------- USERD_set_time_step USERD_get_global_coords USERD_get_block_coords_by_component changing connectivity: --------------------- USERD_set_time_step USERD_get_number_of_model_parts USERD_get_part_build_info USERD_get_number_global_nodes USERD_get_global_coords USERD_get_global_node_ids USERD_get_element_connectivities_for_part USERD_get_element_ids_for_part USERD_get_block_iblanking USERD_get_block_coords_by_component 8. Node or Element queries over time USERD_get_variable_value_at_specific ----------------------- Detailed Specifications ----------------------- Include files: -------------- The following header file is required in any file containing these library routines. #include "global_extern.h" Basis of arrays: --------------- Unless explicitly stated otherwise, all arrays are zero based - in true C fashion. Global variables: ---------------- You will generally need to have a few global variables which are shared by the various library routines. The detailed specifications below have assumed the following are available. (Their names describe their purpose, and they will be used in helping describe the details of the routines below). static int Numparts_available = 0; static int Num_unstructured_parts = 0; static int Num_structured_blocks = 0; /* Note: Numparts_available = Num_unstructured_parts + Num_structured_blocks */ static int Num_time_steps = 1; static int Num_global_nodes = 0; static int Num_variables = 0; static int Num_dataset_files = 0; static int Current_time_step = 0; _________________________________________ ----------------------------------------- Library Routines (in alphabetical order): _________________________________________ ----------------------------------------- -------------------------------------------------------------------- USERD_bkup Description: ----------- This routine is called during the EnSight archive process. You can use it to save or restore info relating to your user defined reader. Specification: ------------- int USERD_bkup(FILE *archive_file, int backup_type) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) archive_file = The archive file pointer (IN) backup_type = Z_SAVE_ARCHIVE for saving archive Z_REST_ARCHIVE for restoring archive Notes: ----- * Since EnSight's archive file is saved in binary form, you should also do any writing to it or reading from it in binary. * You should archive any variables, which will be needed for future operations, that will not be read or computed again before they will be needed. These are typically global variables. * Make sure that the number of bytes that you write on a save and the number of bytes that you read on a restore are identical!! * If any of the variables you save are allocated arrays, you must do the allocations before restoring into them. -------------------------------------------------------------------- USERD_get_block_coords_by_component Description: ----------- Get the coordinates of a given structured block, a component at a time. Specification: ------------- int USERD_get_block_coords_by_component(int block_number, int which_component, float *coord_array) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) block_number = The block part number (IN) which_component = Z_COMPX if x component wanted = Z_COMPY if y component wanted = Z_COMPZ if z component wanted (OUT) coord_array = 1D array containing x,y, or z coordinate component of each node (Array will have been allocated i*j*k for the block long) Notes: ----- * Not called unless Num_structured_blocks is > 0 * Will be based on Current_time_step -------------------------------------------------------------------- USERD_get_block_iblanking Description: ----------- Get the iblanking value at each node of a block (if the block is iblanked). Specification: ------------- int USERD_get_block_iblanking(int block_number, int *iblank_array) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) block_number = The block part number (OUT) iblank_array = 1D array containing iblank value for each node. (Array will have been allocated i*j*k for the block long) possible values are: Z_EXT = exterior Z_INT = interior Z_BND = boundary Z_INTBND = internal boundary Z_SYM = symmetry plane Notes: ----- * Not called unless Num_structured_blocks is > 0 and you have some iblanked blocks * Will be based on Current_time_step -------------------------------------------------------------------- USERD_get_block_scalar_values Description: ----------- if Z_PER_NODE: Get the values at each node of a block, for a given scalar variable or if Z_PER_ELEM: Get the values at each element of a block, for a given scalar variable Specification: ------------- int USERD_get_block_scalar_values(int block_number, int which_scalar, float *scalar_array) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) block_number = The block part number (IN) which_scalar = The variable number (OUT) scalar_array = 1D array containing scalar values for each node or element. Array will have been allocated: if Z_PER_NODE: i*j*k for the block long if Z_PER_ELEM: (i-1)*(i-1)*(k-1) for the block long Notes: ----- * Not called unless Num_structured_blocks is > 0, Num_variables is > 0, and there are some scalar type variables * The per_node or per_elem classification must be obtainable from the variable number (a var_classify array needs to be retained) * Will be based on Current_time_step -------------------------------------------------------------------- USERD_get_block_vector_values_by_component Description: ----------- if Z_PER_NODE: Get the values at each node of a block, for a given vector variable, one component at a time. or if Z_PER_ELEM: Get the values at each element of a block, for a given vector variable, one component at a time. Specification: ------------- int USERD_get_block_vector_values_by_component(int block_number, int which_vector, int which_component, float *vector_array) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) block_number = The block part number (IN) which_vector = The variable number (IN) which_component = Z_COMPX if x component wanted = Z_COMPY if y component wanted = Z_COMPZ if z component wanted (OUT) vector_array = 1D array containing vector component value for each node or element. Array will have been allocated: if Z_PER_NODE: i*j*k for the block long if Z_PER_ELEM: (i-1)*(i-1)*(k-1) for the block long Notes: ----- * Not called unless Num_structured_blocks is > 0, Num_variables is > 0, and there are some vector type variables * The per_node or per_elem classification must be obtainable from the variable number (a var_classify array needs to be retained) * Will be based on Current_time_step -------------------------------------------------------------------- USERD_get_changing_geometry_status Description: ----------- Gets the changing geometry status for the model Specification: ------------- int USERD_get_changing_geometry_status( void ) Returns: ------- Z_STATIC if geometry does not change Z_CHANGE_COORDS if changing coordinates only Z_CHANGE_CONN if changing connectivity Arguments: --------- none Notes: ----- * EnSight does not support changing number of parts. But the coords and/or the connectivity of the parts can change. -------------------------------------------------------------------- USERD_get_constant_value Description: ----------- Get the value of a constant at a time step Specification: ------------- float USERD_get_constant_value(int which_var) Returns: ------- Value of the requested constant variable Arguments: --------- (IN) which_var = The variable number Notes: ----- * Will be based on Current_time_step -------------------------------------------------------------------- USERD_get_dataset_query_file_info Description: ----------- Get the information about files in the dataset. Used for the dataset query option within EnSight. Specification: ------------- int USERD_get_dataset_query_file_info(Z_QFILES *qfiles) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (OUT) qfiles = Structure containing information about each file of the dataset. The Z_QFILES structure is defined in the global_extern.h file (The structure will have been allocated Num_dataset_files long, with 10 description lines per file). qfiles[].name = The name of the file (Z_MAXFILENP is the dimensioned length of the name) qfiles[].sizeb = The number of bytes in the file (Typically obtained with a call to the "stat" system routine) (Is a long) qfiles[].timemod = The time the file was last modified (Z_MAXTIMLEN is the dimensioned length of this string) (Typically obtained with a call to the "stat" system routine) qfiles[].num_d_lines = The number of description lines you are providing from the file. Max = 10 qfiles[].f_desc[] = The description line(s) per file, qfiles[].num_d_lines of them (Z_MAXFILENP is the allocated length of each line) Notes: ----- * If Num_dataset_files is 0, this routine will not be called. -------------------------------------------------------------------- USERD_get_description_lines Description: ----------- Get two description lines associated with geometry per time step, or one description line associated with a variable per time step. Specification: ------------- int USERD_get_description_lines(int which_type, int which_var, char line1[Z_BUFL], char line2[Z_BUFL]) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) which_type = Z_GEOM for geometry (2 lines) = Z_VARI for variable (1 line) (IN) which_var = If it is a variable, which one. Ignored if geometry type. (OUT) line1 = The 1st geometry description line, or the variable description line. (OUT) line2 = The 2nd geometry description line Not used if variable type. Notes: ----- * Will be based on Current_time_step * These are the lines EnSight can echo to the screen in annotation mode. -------------------------------------------------------------------- USERD_get_element_connectivities_for_part Description: ----------- Gets the connectivities for the elements of an unstructured part Specification: ------------- int USERD_get_element_connectivities_for_part(int part_number, int **conn_array[Z_MAXTYPE]) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) part_number = The part number (OUT) conn_array = 3D array containing connectivity of each element of each type. (Array will have been allocated Z_MAXTYPE by num_of_elements of each type by connectivity length of each type) ex) If num_of_elements[Z_TRI03] = 25 num_of_elements[Z_QUA04] = 100 num_of_elements[Z_HEX08] = 30 as obtained in: USERD_get_part_build_info Then the allocated dimensions available for this routine will be: conn_array[Z_TRI03][25][3] conn_array[Z_QUA04][100][4] conn_array[Z_HEX08][30][8] Notes: ----- * Not called unless Num_unstructured_parts is > 0 * Will be based on Current_time_step * The coord_array loaded in USERD_get_global_coords is zero-based, but within EnSight it will become a one-based array. Thus, coord_array[0] will be accessed by node 1 from the conn_array, coord_array[1] will be accessed by node 2 from the conn_array, etc. ex) Given a model of two triangles, you should load coord_array in USERD_get_global_coords as follows: node coordinates ---- ----------- 4 --------- 3 1 coord_array[0].xyz[0] = 0.0 |\ | coord_array[0].xyz[1] = 0.0 | \ T2 | coord_array[0].xyz[2] = 0.0 | \ | | \ | 2 coord_array[1].xyz[0] = 1.0 | \ | coord_array[1].xyz[1] = 0.0 | \ | coord_array[1].xyz[2] = 0.0 | \ | | T1 \ | 3 coord_array[2].xyz[0] = 1.0 | \| coord_array[2].xyz[1] = 1.6 1 --------- 2 coord_array[2].xyz[2] = 0.0 4 coord_array[3].xyz[0] = 0.0 coord_array[3].xyz[1] = 1.6 coord_array[3].xyz[2] = 0.0 And conn_array here as follows: Triangle Connectivity -------- ------------ T1 conn_array[Z_TRI03][0][0] = 1 conn_array[Z_TRI03][0][1] = 2 conn_array[Z_TRI03][0][2] = 4 T2 conn_array[Z_TRI03][1][0] = 2 conn_array[Z_TRI03][1][1] = 3 conn_array[Z_TRI03][1][2] = 4 -------------------------------------------------------------------- USERD_get_element_ids_for_part Description: ----------- Gets the ids for the elements of an unstructured part. Specification: ------------- int USERD_get_element_ids_for_part(int part_number, int *elemid_array[Z_MAXTYPE]) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) part_number = The part number (OUT) elemid_array = 2D array containing id of each element of each type. (Array will have been allocated Z_MAXTYPE by num_of_elements of each type) ex) If num_of_elements[Z_TRI03] = 25 num_of_elements[Z_QUA04] = 100 num_of_elements[Z_HEX08] = 30 as obtained in: USERD_get_part_build_info Then the allocated dimensions available for this routine will be: conn_array[Z_TRI03][25] conn_array[Z_QUA04][100] conn_array[Z_HEX08][30] Notes: ----- * Not called unless Num_unstructured_parts is > 0 and element label status is TRUE * Will be based on Current_time_step -------------------------------------------------------------------- USERD_get_element_label_status Description: ----------- Answers the question as to whether element labels will be provided. Specification: ------------- int USERD_get_element_label_status( void ) Returns: ------- TRUE if element labels will be provided FALSE if element labels will NOT be provided Arguments: --------- none Notes: ----- * element lables are needed in order to do any element querying, or element labeling on-screen within EnSight. For unstructured parts, you can read them from your file if available, or can assign them, etc. They need to be unique per part, and are often unique per model. USERD_get_element_ids_for_part is used to obtain the ids, on a part by part basis, if TRUE status is returned here. For structured parts, EnSight will assign ids if you return a status of TRUE here. You cannot assign them youself!! -------------------------------------------------------------------- USERD_get_global_coords Description: ----------- Gets the coordinates for the global nodes. Specification: ------------- int USERD_get_global_coords(CRD *coord_array) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (OUT) coord_array = 1D array of CRD structures, which contains x,y,z coordinates of each node. (Array will have been allocated Num_global_nodes long) For reference, CRD structure (which is in global_extern) is: typedef struct { float xyz[3]; }CRD; Notes: ----- * Not called unless Num_unstructured_parts is > 0 * Will be based on Current_time_step * The coord_array is zero-based, but within EnSight it will become a one-based array. Thus, coord_array[0] will be accessed by node 1 from the conn_array, coord_array[1] will be accessed by node 2 from the conn_array, etc. ex) Given a model of two triangles, you should load coord_array as follows: node coordinates ---- ----------- 4 --------- 3 1 coord_array[0].xyz[0] = 0.0 |\ | coord_array[0].xyz[1] = 0.0 | \ T2 | coord_array[0].xyz[2] = 0.0 | \ | | \ | 2 coord_array[1].xyz[0] = 1.0 | \ | coord_array[1].xyz[1] = 0.0 | \ | coord_array[1].xyz[2] = 0.0 | \ | | T1 \ | 3 coord_array[2].xyz[0] = 1.0 | \| coord_array[2].xyz[1] = 1.6 1 --------- 2 coord_array[2].xyz[2] = 0.0 4 coord_array[3].xyz[0] = 0.0 coord_array[3].xyz[1] = 1.6 coord_array[3].xyz[2] = 0.0 And conn_array in USERD_get_element_connectivities_for_part as follows: Triangle Connectivity -------- ------------ T1 conn_array[Z_TRI03][0][0] = 1 conn_array[Z_TRI03][0][1] = 2 conn_array[Z_TRI03][0][2] = 4 T2 conn_array[Z_TRI03][1][0] = 2 conn_array[Z_TRI03][1][1] = 3 conn_array[Z_TRI03][1][2] = 4 -------------------------------------------------------------------- USERD_get_global_node_ids Description: ----------- Gets the node ids assigned to each of the global nodes. Specification: ------------- int USERD_get_global_node_ids(int *nodeid_array) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (OUT) nodeid_array = 1D array containing node ids of each node. The ids must be > 0 (Array will have been allocated Num_global_nodes long) Notes: ----- * Not called unless Num_unstructured_parts is > 0 and node label status is TRUE * Will be based on Current_time_step -------------------------------------------------------------------- USERD_get_name_of_reader Description: ----------- Gets the name of your user defined reader. The user interface will ask for this and include it in the available reader list. Specification: ------------- int USERD_get_name_of_reader(char reader_name[Z_MAX_USERD_NAME], int *two_fields) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (OUT) reader_name = the name of the your reader or data format. (max length is Z_MAX_USERD_NAME, which is 20) (OUT) *two_fields = FALSE if only one data field required in the data dialog of EnSight. TRUE if two data fields required. Notes: ----- * Always called. Provide a name for your custom reader format. * If you don't want a custom reader to show up in the data dialog choices, return a name of "No_Custom" -------------------------------------------------------------------- USERD_get_node_label_status Description: ----------- Answers the question as to whether node labels will be provided. Specification: ------------- int USERD_get_node_label_status( void ) Returns: ------- TRUE if node labels will be provided FALSE if node labels will NOT be provided Arguments: --------- none Notes: ----- * Node ids are needed in order to do any node querying, or node labeling on-screen within EnSight. For unstructured parts, you can read them from your file if available, or can assign them, etc. They need to be unique per part, and are often unique per model. The must also be positive numbers greater than zero. USERD_get_global_node_ids is used to obtain the ids, if the status returned here is TRUE. Also be aware that if you say node labels are available, the connectivity of elements must be according to these node ids. For structured parts, EnSight will assign ids if you return a status of TRUE here. You cannot assign them yourself!! -------------------------------------------------------------------- USERD_get_number_of_files_in_dataset Description: ----------- Get the total number of files in the dataset. Used for the dataset query option within EnSight. Specification: ------------- int USERD_get_number_of_files_in_dataset( void ) Returns: ------- The total number of files in the dataset. Arguments: --------- none Notes: ----- * You can be as complete as you want about this. If you don't care about the dataset query option, return a value of 0 If you only want certain files, you can just include them. But, you will need to supply the info in USERD_get_dataset_query_file_info for each file you include here. * Num_dataset_files would be set here -------------------------------------------------------------------- USERD_get_number_of_global_nodes Description: ----------- Gets the number of global nodes, used for unstructured parts. Specification: ------------- int USERD_get_number_of_global_nodes() Returns: ------- Number of global nodes (>=0 if okay, <0 if problems) Arguments: --------- none Notes: ----- * Not called unless Num_unstructured_parts is > 0 * Will be based on Current_time_step * For unstructured data: EnSight wants 1. A global array of nodes 2. Element connectivities by part, which reference the node numbers of the global node array. IMPORTANT: --------- If you provide node ids, then element connectivities must be in terms of the node ids. If you do not provide node ids, then element connectivities must be in terms of the index into the node array, but shifted to start at 1 * Num_global_nodes would be set here -------------------------------------------------------------------- USERD_get_number_of_model_parts Description: ----------- Gets the total number of unstructured and structured parts in the model, for which you can supply information. Specification: ------------- int USERD_get_number_of_model_parts( void ) Returns: ------- Number of parts (>0 if okay, <=0 if probs). Arguments: --------- none Notes: ----- * If going to have to read down through the parts in order to know how many, you may want to build a table of pointers to the various parts, so you can easily get to particular parts in later processes. If you can simply read the number of parts at the head of the file, then you would probably not build the table at this time. * This routine would set Numparts_available, which is equal to Num_unstructured_parts + Num_structured_blocks. -------------------------------------------------------------------- USERD_get_number_of_time_steps Description: ----------- Gets the number of time steps of data available. Specification: ------------- int USERD_get_number_of_time_steps( void ) Returns: ------- Number of time steps (>0 if okay, <=0 if problems). Arguments: --------- none Notes: ----- * This should be >= 1 1 indicates a static model >1 indicates a transient model * Num_time_steps would be set here -------------------------------------------------------------------- USERD_get_number_of_variables Description: ----------- Get the number of variables for which you will be providing info. Specification: ------------- int USERD_get_number_of_variables( void ) Returns: ------- Number of variables (includes constant, scalar, and vector types) (>=0 if okay, <0 if problem) Arguments: --------- none Notes: ----- ***************************************************************** * Variable numbers, by which references will be made, are implied here. If you say there are 3 variables, the variable numbers will be 1, 2, and 3. ***************************************************************** * Num_variables would be set here -------------------------------------------------------------------- USERD_get_part_build_info Description: ----------- Gets the info needed for the part building process. Specification: ------------- int USERD_get_part_build_info(int *part_numbers, int *part_types, char *part_description[Z_BUFL], int *number_of_elements[Z_MAXTYPE], int *ijk_dimensions[3], int *iblanking_options[6]) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (OUT) part_numbers = Array containing part numbers for each of the model parts. IMPORTANT: Parts numbers must be >= 1 ******************************************** The numbers provided here are the ones by which the parts will be referred to in any of the other routines which receive a part number or block number as an argument!! ******************************************** (Array will have been allocated Numparts_available long) (OUT) part_types = Array containing one of the following for each model part: Z_UNSTRUCTURED or Z_STRUCTURED or Z_IBLANKED (Array will have been allocated Numparts_available long) (OUT) part_description = Array containing a description for each of the model parts (Array will have been allocated Numparts_available by Z_BUFL long) (OUT) number_of_elements = 2D array containing number of each type of element for each unstructured model part. ------------ Possible types are: Z_POINT = point Z_BAR02 = 2-noded bar Z_BAR03 = 3-noded bar Z_TRI03 = 3-noded triangle Z_TRI06 = 6-noded triangle Z_QUA04 = 4-noded quadrilateral Z_QUA08 = 8-noded quadrilateral Z_TET04 = 4-noded tetrahedron Z_TET10 = 10-noded tetrahedron Z_PYR05 = 5-noded pyramid Z_PYR13 = 13-noded pyramid Z_PEN06 = 6-noded pentahedron Z_PEN15 = 15-noded pentahedron Z_HEX08 = 8-noded hexahedron Z_HEX20 = 20-noded hexahedron (Ignored unless Z_UNSTRUCTURED type) (Array will have been allocated Numparts_available by Z_MAXTYPE long) (OUT) ijk_dimensions = 2D array containing ijk dimensions for each structured model part. ---------- (Ignored if Z_UNSTRUCTURED type) (Array will have been allocated Numparts_available by 3 long) ijk_dimensions[][0] = I dimension ijk_dimensions[][1] = J dimension ijk_dimensions[][2] = K dimension (OUT) iblanking_options = 2D array containing iblanking options possible for each structured model part. ---------- (Ignored unless Z_IBLANKED type) (Array will have been allocated Numparts_available by 6 long) iblanking_options[][Z_EXT] = TRUE if external (outside) [][Z_INT] = TRUE if internal (inside) [][Z_BND] = TRUE if boundary [][Z_INTBND] = TRUE if internal boundary [][Z_SYM] = TRUE if symmetry surface Notes: ----- * If you haven't built a table of pointers to the different parts, you might want to do so here as you gather the needed info. * Will be based on Current_time_step -------------------------------------------------------------------- USERD_get_scalar_values Description: ----------- if Z_PER_NODE: Get the values at each global node for a given scalar variable. or if Z_PER_ELEM: Get the values at each element of a specific part and type for a given scalar variable. Specification: ------------- int USERD_get_scalar_values(int which_scalar, int which_part, int which_type, float *scalar_array) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) which_scalar = The variable number (of scalar type) (IN) which_part if Z_PER_NODE: Not used if Z_PER_ELEM: = The part number (IN) which_type if Z_PER_NODE: Not used if Z_PER_ELEM: = The element type (OUT) scalar_array if Z_PER_NODE: = 1D array containing scalar values for each node. (Array will have been allocated Num_global_nodes long) if Z_PER_ELEM: = 1d array containing scalar values for each element of a particular part and type. (Array will have been allocated number_of_elements[which_part][which_type] long. See USERD_get_part_build_info) Notes: ----- * Not called unless Num_unstructured_parts is > 0, Num_variables is > 0, and you have some scalar type variables. * The per_node or per_elem classification must be obtainable from the variable number (a var_classify array needs to be retained) * Will be based on Current_time_step -------------------------------------------------------------------- USERD_get_solution_times Description: ----------- Get the solution times associated with each time step. Specification: ------------- int USERD_get_solution_times(float *solution_times) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (OUT) solution_times = 1D array of solution times/time step (Array will have been allocated Num_time_steps long) Notes: ----- * The solution times must be non-negative and increasing. -------------------------------------------------------------------- USERD_get_variable_info Description: ----------- Get the variable descriptions, types and filenames Specification: ------------- int USERD_get_variable_info(char **var_description, char **var_filename, int *var_type, int *var_classify) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (OUT) var_description = Variable descriptions (Array will have been allocated Num_variables by Z_BUFL long) (OUT) var_filename = Variable filenames (Array will have been allocated Num_variables by Z_BUFL long) (OUT) var_type = Variable type (Array will have been allocated Num_variables long) types are: Z_CONSTANT Z_SCALAR Z_VECTOR (OUT) var_classify = Variable classification (Array will have been allocated Num_variables long) types are: Z_PER_NODE Z_PER_ELEM Notes: ----- * The implied variable numbers apply, but be aware that the arrays are zero based. So for variable 1, will need to provide var_description[0] var_filename[0] var_type[0] var_classify[0] for variable 2, will need to provide var_description[1] var_filename[1] var_type[1] var_classify[1] etc. -------------------------------------------------------------------- USERD_get_variable_value_at_specific Description: ----------- if Z_PER_NODE: Get the value of a particular variable at a particular node in a particular part at a particular time. or if Z_PER_ELEM: Get the value of a particular variable at a particular element of a particular type in a particular part at a particular time. Specification: ------------- int USERD_get_variable_value_at_specific(int which_var, int which_node_or_elem, int which_part, int which_elem_type, int time_step, float values[3]) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) which_var = The variable number (IN) which_node_or_elem If Z_PER_NODE: = The node number. This is not the id, but is the index of the global node list (1 based), or the block's node list (1 based). Thus, coord_array[1] coord_array[2] coord_array[3] . | . |which_node_or_elem index . ---- If Z_PER_ELEM: = The element number. This is not the id, but is the element number index of the number_of_element array (see USERD_get_part_build_info), or the block's element list (1 based). Thus, for which_part: conn_array[which_elem_type][0] conn_array[which_elem_type][1] conn_array[which_elem_type][2] . | . which_node_or_elem index . ---- (IN) which_part If Z_PER_NODE, or block part: = Not used If Z_PER_ELEM: = The part number (IN) which_elem_type If Z_PER_NODE, or block part: = Not used If Z_PER_ELEM: = The element type. This is the element type index of the number_of_element array (see USERD_get_part_build_info) (IN) time_step = The time step (OUT) values = scalar or vector component value(s) values[0] = scalar or vector[0] values[1] = vector[1] values[2] = vector[2] Notes: ----- * This routine is used in node querys over time (or element querys over time for Z_PER_ELEM variables). If these operations are not critical to you, this can be a dummy routine. * The per_node or per_elem classification must be obtainable from the variable number (a var_classify array needs to be retained) -------------------------------------------------------------------- USERD_get_vector_values Description: ----------- if Z_PER_NODE: Get the values at each global node for a given vector variable. or if Z_PER_ELEM: Get the values at each element of a specific part and type for a given vector variable. Specification: ------------- int USERD_get_vector_values(int which_vector, int which_part, int which_type, float *vector_array) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) which_vector = The variable number (IN) which_part if Z_PER_NODE: Not used if Z_PER_ELEM: = The part number (IN) which_type if Z_PER_NODE: Not used if Z_PER_ELEM: = The element type (OUT) vector_array if Z_PER_NODE: = 1D array containing vector values for each node. (Array will have been allocated 3 by Num_global_nodes long) Info stored in this fashion: vector_array[0] = xcomp of node 1 vector_array[1] = ycomp of node 1 vector_array[2] = zcomp of node 1 vector_array[3] = xcomp of node 2 vector_array[4] = ycomp of node 2 vector_array[5] = zcomp of node 2 vector_array[6] = xcomp of node 3 vector_array[7] = ycomp of node 3 vector_array[8] = zcomp of node 3 etc. if Z_PER_ELEM: = 1D array containing vector values for each element of a particular part and type. (Array will have been allocated 3 by number_of_elements[which_part][which_type] long. See USERD_get_part_build_info) Info stored in this fashion: vector_array[0] = xcomp of elem 1 (of part and type) vector_array[1] = ycomp of elem 1 " vector_array[2] = zcomp of elem 1 " vector_array[3] = xcomp of elem 2 " vector_array[4] = ycomp of elem 2 " vector_array[5] = zcomp of elem 2 " vector_array[6] = xcomp of elem 3 " vector_array[7] = ycomp of elem 3 " vector_array[8] = zcomp of elem 3 " etc. Notes: ----- * Not called unless Num_unstructured_parts is > 0, Num_variables is > 0, and you have some vector type variables * The per_node or per_elem classification must be obtainable from the variable number (a var_classify array needs to be retained) * Will be based on Current_time_step -------------------------------------------------------------------- USERD_set_filenames Description: ----------- Receives the geometry and result filenames entered in the data dialog. The user written code will have to store and use these as needed. Specification: ------------- int USERD_set_filenames(char filename_1[], char filename_2[], char the_path[], int swapbytes) Returns: ------- Z_OK if successful Z_ERR if not successful Arguments: --------- (IN) filename_1 = the filename entered into the geometry field of the data dialog. (IN) filename_2 = the filename entered into the result field of the data dialog. (If the two_fields flag in USERD_get_name_of_reader is FALSE, this will be null string) (IN) the_path = the path info from the data dialog. Note: filename_1 and filename_2 have already had the path prepended to them. This is provided in case it is needed for filenames contained in one of the files (IN) swapbytes = TRUE if should swap bytes when reading data. Notes: ----- * Since you must manage everything from the input that is entered in these data dialog fields, this is an important routine! * It may be that you will need to have an executive type file that contains info and other filenames within it, like EnSight6's case file. -------------------------------------------------------------------- USERD_set_time_step Description: ----------- Set the current time step. All functions that need time, and that do not explicitly pass it in, will use the time step set by this routine. Specification: ------------- void USERD_set_time_step(int time_step) Returns: ------- nothing Arguments: --------- (IN) time_step - The current time step to set Notes: ----- * Current_time_step would be set here -------------------------------------------------------------------- USERD_stop_part_building Description: ----------- This routine called when the part building dialog is closed. It is provided in case you desire to release memory, etc. that was only needed during the part building process. Specification: ------------- void USERD_stop_part_building( void ) Returns: ------- nothing Arguments: --------- none Notes: ----- ---- end of doucment ----