The CGNS Mid-Level Library may be downloaded from the CGNS site . The manual, as well as the other CGNS documentation, is available from the CGNS documentation web site.
The sections that follow describe the Mid-Level Library functions in detail. The first three sections cover some basic file operations (i.e., opening and closing a CGNS file, and some configuration options), accessing a specific node in a CGNS database, and error handling. The remaining sections describe the functions used to read, write, and modify nodes and data in a CGNS database. These sections basically follow the organization used in the "Detailed CGNS Node Descriptions" section of the SIDS File Mapping manual.
At the start of each sub-section is a Node line, listing the the applicable CGNS node label, with links to that node's description in the SIDS and SIDS File Mapping manuals.
Next is a table illustrating the syntax for the Mid-Level Library functions. The C functions are shown in the top half of the table, followed by the corresponding Fortran routines in the bottom half of the table. Fortran subroutines identified in green indicates APIs which do not have explicit interfaces. Input variables are shown in an upright blue font, and output variables are shown in a slanted red font. Note, as of CGNS-3.1.0, some of the arguments to the Mid-Level Library have changed from int to cgsize_t in order to support 64-bit data. Changed APIs can quickly be identified by searching for cgsize_t. For each function, the right-hand column lists the modes (read, write, and/or modify) applicable to that function.
The input and output variables are then listed and defined.
The CGNS Mid-Level Library is written in C, but each function has a Fortran counterpart. All function names start with "cg_". The Fortran functions have the same name as their C counterpart with the addition of the suffix "_f".
All data structure names and labels in CGNS are limited to 32 characters. When reading a file, it is advised to pre-allocate the character string variables to 32 characters in Fortran, and 33 in C (to include the string terminator). Other character strings, such as the CGNS file name or descriptor text, are unlimited in length. The space for unlimited length character strings will be created by the Mid-Level Library; it is then the responsibility of the application to release this space by a call to cg_free.
All C functions return an integer value representing the error status. All Fortran functions have an additional parameter, ier, which contains the value of the error status. An error status different from zero implies that an error occured. The error message can be printed using the error handling functions of the CGNS library. The error codes are coded in the C and Fortran include files cgnslib.h and cgnslib_f.h.
Beginning with CGNS-3.1.0, two new typedef variables have been introduced to support 64-bit mode. The cglong_t typedef is always a 64-bit integer, and cgsize_t will be either a 32-bit or 64-bit integer depending on how the library was built. Many of the C functions in the MLL have been changed to to use cgsize_t instead of int in the arguments. These functions include any that may exceed the 2Gb limit of an int, e.g. zone dimensions, element data, boundary conditions, and connectivity. In Fortran, all integer data is taken to be integer*4 for 32-bit and integer*8 for 64-bit builds.
Several types of variables are defined using typedefs in the cgnslib.h file. These are intended to facilitate the implementation of CGNS in C. These variable types are defined as an enumeration of key words admissible for any variable of these types. The file cgnslib.h must be included in any C application programs which use these data types.
In Fortran, the same key words are defined as integer parameters in the include file cgnslib_f.h. Such variables should be declared as integer in Fortran applications. The file cgnslib_f.h must be included in any Fortran application using these key words.
Note that the first two enumerated values in these lists, xxxNull and xxxUserDefined, are only available in the C interface, and are provided in the advent that your C compiler does strict type checking. In Fortran, these values are replaced by the numerically equivalent CG_Null and CG_UserDefined. These values are also defined in the C interface, thus either form may be used. The function prototypes for the MLL use CG_Null and CG_UserDefined, rather than the more specific values.
The list of enumerated values (key words) for each of these variable
types (typedefs) are:
|ZoneType_t||ZoneTypeNull, ZoneTypeUserDefined, Structured, Unstructured|
|ElementType_t||ElementTypeNull, ElementTypeUserDefined, NODE, BAR_2, BAR_3, TRI_3, TRI_6, QUAD_4, QUAD_8, QUAD_9, TETRA_4, TETRA_10, PYRA_5, PYRA_14, PENTA_6, PENTA_15, PENTA_18, HEXA_8, HEXA_20, HEXA_27, MIXED, PYRA_13, NGON_n, NFACE_n, BAR_4, TRI_9, TRI_10, QUAD_12, QUAD_16, TETRA_16, TETRA_20, PYRA_21, PYRA_29, PYRA_30, PENTA_24, PENTA_38, PENTA_40, HEXA_32, HEXA_56, HEXA_64, BAR_5, TRI_12, TRI_15, QUAD_P4_16, QUAD_25, TETRA_22, TETRA_34, TETRA_35, PYRA_P4_29, PYRA_50, PYRA_55, PENTA_33, PENTA_66, PENTA_75, HEXA_44, HEXA_98, HEXA_125|
Integer, RealSingle, RealDouble, Character, LongInteger
Kilogram, Gram, Slug, PoundMass
Meter, Centimeter, Millimeter, Foot, Inch
Kelvin, Celsius, Rankine, Fahrenheit
Ampere, Abampere, Statampere, Edison, auCurrent
Mole, Entities, StandardCubicFoot, StandardCubicMeter
Candela, Candle, Carcel, Hefner, Violle
FullPotential, Euler, NSLaminar,
Ideal, VanderWaals, Constant,
PowerLaw, SutherlandLaw, ConstantPrandtl, EddyViscosity,
Frozen, ThermalEquilib, ThermalNonequilib,
ChemicalNonequilib, EMElectricField, EMMagneticField,
EMConductivity, Voltage, Interpolated,
Vertex, CellCenter, FaceCenter, IFaceCenter, JFaceCenter,
Overset, Abutting, Abutting1to1
PointList, PointRange, PointListDonor, PointRangeDonor,
ElementList, ElementRange, CellListDonor
BCDegeneratePoint, BCDirichlet, BCFarfield, BCNeumann,
BCGeneral, BCInflow, BCOutflow, BCInflowSubsonic,
BCOutflowSupersonic, BCSymmetryPlane, BCTunnelInflow,
BCSymmetryPolar, BCTunnelOutflow, BCWallViscous,
BCWall, BCWallViscousHeatFlux, BCWallInviscid,
AverageAll, AverageCircumferential, AverageRadial, AverageI,
The CGNS library defines character arrays which map the typedefs above to character strings. These are global arrays dimensioned to the size of each list of typedefs. To retrieve a character string representation of a typedef, use the typedef value as an index to the appropiate character array. For example, to retrieve the string "Meter" for the LengthUnits_t Meter typedef, use LengthUnitsName[Meter]. Functions are available to retrieve these names without the need for direct global data access. These functions also do bounds checking on the input, and if out of range, will return the string "<invalid>". An additional benefit is that these will work from within a Windows DLL, and are thus the recommended access technique. The routines have the same name as the global data arrays, but with a "cg_" prepended. For the example above, use "cg_LengthUnitsName(Meter)".
|Typedef Name Access Functions|
|const char *name = cg_MassUnitsName(MassUnits_t type);|
|const char *name = cg_LengthUnitsName(LengthUnits_t type);|
|const char *name = cg_TimeUnitsName(TimeUnits_t type);|
|const char *name = cg_TemperatureUnitsName(TemperatureUnits_t type);|
|const char *name = cg_ElectricCurrentUnitsName(ElectricCurrentUnits_t type);|
|const char *name = cg_SubstanceAmountUnitsName(SubstanceAmountUnits_t type);|
|const char *name = cg_LuminousIntensityUnitsName(LuminousIntensityUnits_t type);|
|const char *name = cg_DataClassName(DataClass_t type);|
|const char *name = cg_GridLocationName(GridLocation_t type);|
|const char *name = cg_BCDataTypeName(BCDataType_t type);|
|const char *name = cg_GridConnectivityTypeName(GridConnectivityType_t type);|
|const char *name = cg_PointSetTypeName(PointSetType_t type);|
|const char *name = cg_GoverningEquationsTypeName(GoverningEquationsType_t type);|
|const char *name = cg_ModelTypeName(ModelType_t type);|
|const char *name = cg_BCTypeName(BCType_t type);|
|const char *name = cg_DataTypeName(DataType_t type);|
|const char *name = cg_ElementTypeName(ElementType_t type);|
|const char *name = cg_ZoneTypeName(ZoneType_t type);|
|const char *name = cg_RigidGridMotionTypeName(RigidGridMotionType_t type);|
|const char *name = cg_ArbitraryGridMotionTypeName(ArbitraryGridMotionType_t type);|
|const char *name = cg_SimulationTypeName(SimulationType_t type);|
|const char *name = cg_WallFunctionTypeName(WallFunctionType_t type);|
|const char *name = cg_AreaTypeName(AreaType_t type);|
|const char *name = cg_AverageInterfaceTypeName(AverageInterfaceType_t type);|
If you use the cgsize_t data type in new code, it will work in both 32 and 64-bit compilation modes. In order to support CGNS versions prior to 3.1, you may also want to add something like this to your code:
#if CGNS_VERSION < 3100 #define cgsize_t int #endifExisting code that uses int will not work with a CGNS 3.1 library compiled in 64-bit mode. You may want to add something like this to your code:
#if CGNS_VERSION >= 3100 && CG_BUILD_64BIT #error does not work in 64 bit mode #endifor modify your code to use cgsize_t.
Starting with CGNS-3.3.0, a new CGNS module was added to the library. Fortran programs can use the new module by adding USE CGNS. The use of include 'cgnslib_f.h' is deprecated as of CGNS-3.3.0.
|Fortran Helper Functions||Modes|
Returns the data type of buf, where buf is a scalar. This is a useful function for automatically passing the correct data type of a buffer.
For example:CALL cg_coord_read_f(cg, base, zone, coordname, cg_get_type(data(1)), rmin, DataSize, data, ier)
Fortran APIs which can except a null character or an empty string are encouraged to pass C_NULL_CHAR as opposed to "\0" or "".
Starting with CGNS-3.3.0, the Fortran APIs have the following specifications (recommended for portability):
An integer parameter, CG_BUILD_64BIT, can be used to tell the size of cgsize_t, which will be set to 1 in 64-bit mode and 0 otherwise. You may use this parameter to check at run time if the CGNS library has been compiled in 64-bit mode or not, as in:
if (CG_BUILD_64BIT .ne. 0) then print *,'will not work in 64-bit mode' stop endifIf you are using a CGNS library prior to version 3.1, this parameter will not be defined and you will need to rely on your compiler initializing all undefined values to 0 (not always the case) for this test to work.
DISCLAMER: The following section's practice is not recommend!
If you have explicitly defined your default integers which are passed to the CGNS library as INTEGER*8, or used a compiler option to promote implicit integers to INTEGER*8, then you MUST compile the CGNS library with the same compiler option in order to promote implicit integers to INTEGER*8. If you really must promote all integers to INTEGER*8 in your code, and you are not able to compile the CGNS library with the same compilar options, then it is recommended that all arguments in the CGNS Fortran APIs should be declared as INTEGER(C_INT) if the corresponding argument in the C API is declared as an int.