VisTags.cpp

tool for visualizing multi level tags
To run: VisTags <inp_file> <outfile> -O <read_opts> -t <tags> -l <levels> -d <dim>

In this example, it is shown how to create some simple tags for those tags that come from climate data, multiple levels. you can read directly nc data, or *.h5m file that will have the tag with multi levels output will be a vtk file with dense tags of form tag_name_<level> the tag name might contain a time index too, like T0 or U0 <tag> is a list of tags, separated by commas, no spaces <levels> is a list of levels, separated by commas, no spaces dimension of entities with the tags will be specified with -d (default 2)

an example of use

VisTags gcrm_r3.nc out.vtk -O VARIABLE=u -t u0,u1 -l 0,1,2 -d 2 (we knew that it had variable u in the file, that it had 256 levels, that there are 2 time steps, etc)

or VisTags gcrm_r3.nc out.vtk -t u0 -l 0,1,2 -d 2 (it will read all variables, but we need to know that u0 will be created as a tag)

the out.vtk file will contain u0_0, u0_1, as simple dense double tags

/** @example VisTags.cpp \n
 * \brief tool for visualizing multi level tags  \n
 * <b>To run</b>: VisTags  <inp_file>  <outfile> -O <read_opts> -t <tags> -l <levels>  -d <dim> \n
 *
 * In this example, it is shown how to create some simple tags for those tags that come from
 *  climate data, multiple levels.
 *  you can read directly nc data, or *.h5m file that will have the tag with multi levels
 *   output will be a vtk file with dense tags of form tag_name_<level>
 * the tag name might contain a time index too, like T0 or U0
 * <tag> is a list of tags, separated by commas, no spaces
 * <levels> is a list of levels, separated by commas, no spaces
 *  dimension of entities with the tags will be specified with -d (default 2)
 *
 * an example of use
 *
 * VisTags gcrm_r3.nc out.vtk -O VARIABLE=u -t u0,u1 -l 0,1,2 -d 2
 * (we knew that it had variable u in the file, that it had 256 levels, that there are 2 time
 *  steps, etc)
 *
 * or
 *  VisTags gcrm_r3.nc out.vtk  -t u0 -l 0,1,2 -d 2
 *  (it will read all variables, but we need to know that u0 will be created as a tag)
 *
 *  the out.vtk file will contain u0_0, u0_1, as simple dense double tags
 */

#include <iostream>
#include <vector>
#include <sstream>
#include <string>

// Include header for MOAB instance and tag conventions
#include "moab/Core.hpp"
#include "MBTagConventions.hpp"
#include "moab/FileOptions.hpp"
#ifdef MOAB_HAVE_MPI
#include "moab/ParallelComm.hpp"
#endif
using namespace moab;
using namespace std;

int main( int argc, char** argv )
{
#ifdef MOAB_HAVE_NETCDF

#ifdef MOAB_HAVE_MPI
    MPI_Init( &argc, &argv );
#endif

    ErrorCode rval;
    string file_input, file_output;
    string read_opts, tags;  // Tags to write, separated by commas; it is the name of the tag
    if( argc < 2 )
    {
        file_input  = string( MESH_DIR ) + string( "/io/gcrm_r3.nc" );
        file_output = "VisTagsOut.vtk";
    }
    else
    {
        file_input  = argv[1];
        file_output = argv[2];
    }
    read_opts = "";
    tags      = "";

    // Instantiate
    Interface* mb = new( std::nothrow ) Core;
    if( NULL == mb ) return 1;

    int dimension = 2;
    // In MOAB, it may have index after reading (T0, T1, etc)
    char* levels = NULL;  // Levels, separated by commas, no spaces (like 0, 1, 19)
    if( argc > 3 )
    {
        int index = 3;
        while( index < argc )
        {
            if( !strcmp( argv[index], "-O" ) )  // This is for reading options, optional
                read_opts = argv[++index];
            if( !strcmp( argv[index], "-t" ) ) tags = argv[++index];
            if( !strcmp( argv[index], "-l" ) ) levels = argv[++index];
            if( !strcmp( argv[index], "-d" ) ) dimension = atoi( argv[++index] );
            index++;
        }
    }

    ostringstream opts;
    opts << ";;TAGS=" << tags << ";LEVELS=" << levels << "\0";
    FileOptions fo( opts.str().c_str() );

    vector< string > tagsNames;
    vector< int > levelsArray;
    fo.get_strs_option( "TAGS", tagsNames );
    fo.get_ints_option( "LEVELS", levelsArray );

    // Load the input file with the specified options
    rval = mb->load_file( file_input.c_str(), 0, read_opts.c_str() );MB_CHK_SET_ERR( rval, "not loading file" );

    Range ents;
    rval = mb->get_entities_by_dimension( 0, dimension, ents );MB_CHK_SET_ERR( rval, "not getting ents" );

    // Now create double tags for entities of dimension
    for( size_t i = 0; i < tagsNames.size(); i++ )
    {
        string tagName = tagsNames[i];
        Tag tagh;
        rval = mb->tag_get_handle( tagName.c_str(), tagh );
        if( MB_SUCCESS != rval )
        {
            cout << "not getting tag " << tagName.c_str() << "\n";
            continue;
        }

        int len = 0;
        rval    = mb->tag_get_length( tagh, len );
        if( MB_SUCCESS != rval )
        {
            cout << "not getting tag len " << tagName.c_str() << "\n";
            continue;
        }

        DataType type;
        rval = mb->tag_get_data_type( tagh, type );
        if( MB_SUCCESS != rval )
        {
            cout << "not getting tag type " << tagName.c_str() << "\n";
            continue;
        }

        int count;
        void* dataptr;  // Assume double tags, for simplicity
        rval = mb->tag_iterate( tagh, ents.begin(), ents.end(), count, dataptr );
        if( MB_SUCCESS != rval || count != (int)ents.size() )
        {
            cout << "not getting tag iterate right " << tagName.c_str() << "\n";
            continue;
        }

        // Now create a new tag, with a new name, concatenated, and copy data there , for each level
        for( size_t j = 0; j < levelsArray.size(); j++ )
        {
            int level = levelsArray[j];
            if( level >= len )
            {
                cout << "level too big at " << level << "\n";
                continue;
            }

            ostringstream newTagName;
            newTagName << tagName << "_" << level;
            Tag newTagh;
            rval = mb->tag_get_handle( newTagName.str().c_str(), 1, type, newTagh, MB_TAG_DENSE | MB_TAG_CREAT );
            if( MB_SUCCESS != rval )
            {
                cout << "not getting new tag " << newTagName.str() << "\n";
                continue;
            }

            void* newDataPtr;
            rval = mb->tag_iterate( newTagh, ents.begin(), ents.end(), count, newDataPtr );
            if( MB_SUCCESS != rval || count != (int)ents.size() )
            {
                cout << "not getting new tag iterate " << newTagName.str() << "\n";
                continue;
            }

            if( MB_TYPE_DOUBLE == type )
            {
                double* ptrD    = (double*)newDataPtr;
                double* oldData = (double*)dataptr;
                for( int k = 0; k < count; k++, ptrD++ )
                    *ptrD = oldData[level + count * k];
            }
        }  // for (size_t j = 0; j < levelsArray.size(); j++)

        mb->tag_delete( tagh );  // No need for the tag anymore, write it to the new file
    }                            // for (size_t i = 0; i < tagsNames.size(); i++)

    rval = mb->write_file( file_output.c_str() );MB_CHK_SET_ERR( rval, "Can't write file " << file_output );
    cout << "Successfully wrote file " << file_output << "\n";

    delete mb;

#ifdef MOAB_HAVE_MPI
    MPI_Finalize();
#endif
#else
    std::cout << " configure with netcdf for this example to work\n";
#endif
    return 0;
}