Index Page
A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X 



   void ekifld_c ( SpiceInt           handle,
                   ConstSpiceChar   * tabnam,
                   SpiceInt           ncols,
                   SpiceInt           nrows,
                   SpiceInt           cnmlen,
                   const void       * cnames,
                   SpiceInt           declen,
                   const void       * decls,
                   SpiceInt         * segno,
                   SpiceInt         * rcptrs )


   Initialize a new E-kernel segment to allow fast writing. 






   Variable  I/O  Description 
   --------  ---  -------------------------------------------------- 
   handle     I   File handle. 
   tabnam     I   Table name. 
   ncols      I   Number of columns in the segment. 
   nrows      I   Number of rows in the segment. 
   cnmlen     I   Length of names in in column name array.
   cnames     I   Names of columns. 
   declen     I   Length of declaration strings in declaration array.
   decls      I   Declarations of columns. 
   segno      O   Segment number. 
   rcptrs     O   Array of record pointers. 


   handle         is the handle of an EK file open for write access. 
                  A new segment is to be created in this file. 
   tabnam         is the name of the EK table to which the current
                  segment belongs.  All segments in the EK file
                  designated by handle must have identical column
                  attributes. tabnam must not exceed SPICE_EK_TNAMSZ
                  (see SpiceEK.h) characters in length.  Case is not
                  significant.  Table names must start with a letter
                  and contain only characters from the set
   ncols          is the number of columns in a new segment. 
   nrows          is the number of rows in a new segment.  Each 
                  column to be added to the segment must contain 
                  the number of entries indicated by nrows. 
   cnames         are, respectively, the length of the column name
                  strings in the column name array, and the base
                  address of the array itself.  The array should have
   decls          are, respectively, the length of the declaration
                  strings in the declaration array, and the base 
                  address of the array itself.  The array should have
                  The Ith element of cnames and the Ith element of decls
                  apply to the Ith column in the segment. 
                  Column names must not exceed SPICE_EK_CNAMSZ (see
                  SpiceEK.h) characters in length.  Case is not
                  significant.  Column names must start with a letter
                  and contain only characters from the set
                  The declarations are strings that contain 
                  `keyword=value' assignments that define the 
                  attributes of the columns to which they apply.  The 
                  column attributes that are defined by a column 
                  declaration are: 
                     <is the column indexed?> 
                     <does the column allow null values?> 
                  The form of a declaration is 
                     "DATATYPE  = <type>, 
                      SIZE      = <size>, 
                      INDEXED   = <boolean>, 
                      NULLS_OK  = <boolean>" 
                  For example, an indexed, scalar, integer column 
                  that allows null values would have the declaration 
                     "DATATYPE  = INTEGER, 
                      SIZE      = 1, 
                      INDEXED   = TRUE, 
                      NULLS_OK  = TRUE" 
                  Commas are required to separate the assignments 
                  within declarations; white space is optional; 
                  case is not significant. 
                  The order in which the attribute keywords are 
                  listed in declaration is not significant. 
                  Every column in a segment must be declared. 
                  Each column entry is effectively an array, each 
                  element of which has the declared data type.  The 
                  SIZE keyword indicates how many elements are in 
                  each entry of the column in whose declaration the 
                  keyword appears.  Note that only scalar-valued 
                  columns (those for which SIZE = 1) may be 
                  referenced in query constraints.  A size 
                  assignment has the syntax 
                     SIZE = <integer> 
                     SIZE = VARIABLE 
                  The size value defaults to 1 if omitted. 
                  The DATATYPE keyword defines the data type of 
                  column entries.  The DATATYPE assignment syntax 
                  has any of the forms 
                     DATATYPE = CHARACTER*(<length>) 
                     DATATYPE = CHARACTER*(*) 
                     DATATYPE = DOUBLE PRECISION 
                     DATATYPE = INTEGER 
                     DATATYPE = TIME 
                  As the datatype declaration syntax suggests, 
                  character strings may have fixed or variable 
                  length.  Variable-length strings are allowed only 
                  in columns of size 1. 
                  Optionally, scalar-valued columns may be indexed. 
                  To create an index for a column, use the assignment 
                     INDEXED = TRUE 
                  By default, columns are not indexed. 
                  Optionally, any column can allow null values.  To 
                  indicate that a column may allow null values, use 
                  the assigment 
                     NULLS_OK = TRUE 
                  in the column declaration.  By default, null 
                  values are not allowed in column entries. 


   segno          is the number of the segment to which data is to be
                  added. Segments are numbered from 0 to nseg-1, where
                  nseg is the count of segments in the file.  Segment
                  numbers are used as unique identifiers by other EK
                  access routines.
   rcptrs         is an array of record pointers for the input 
                  segment.  This array must not be modified by the 
                  The array rcptrs must be passed as an input to 
                  each column addition routine called while 
                  writing the specified segment. 
                  rcptrs must be declared with dimension nrows. 




   1)  If HANDLE is invalid, the error will be diagnosed by routines 
       called by this routine. 
   2)  If TABNAM is more than SPICE_EK_TNAMSZ characters long, the
       error is diagnosed by routines called by this routine.
   3)  If TABNAM contains any nonprintable characters, the error 
       is diagnosed by routines called by this routine. 
   4)  If NCOLS is non-positive, the error is diagnosed by routines 
       called by this routine. 
   5)  If NROWS is non-positive, the error SPICE(INVALIDCOUNT) 
       is signalled. 
   6)  If any column name exceeds SPICE_EK_CNAMSZ characters in length,
       the error is diagnosed by routines called by this routine.
   7)  If any column name contains non-printable characters, the 
       error is diagnosed by routines called by this routine. 
   8)  If a declaration cannot be understood by this routine, the 
       error is diagnosed by routines called by this routine. 
   9)  If an non-positive string length or element size is specified, 
       the error is diagnosed by routines called by this routine. 
  10)  If an I/O error occurs while reading or writing the indicated 
       file, the error will be diagnosed by routines called by this 


   See the EK Required Reading for a discussion of the EK file 


   This routine prepares an EK for the creation of a new segment via 
   the fast column writer routines.  After this routine is called, 
   the columns of the segment are filled in by calls to the fast 
   column writer routines of the appropriate data types.  The fast 
   column writer routines are: 
      ekaclc_c {EK, add column, character} 
      ekacld_c {EK, add column, double precision} 
      ekacli_c {EK, add column, integer} 
   When all of the columns have been added, the write operation is 
   completed by a call to ekffld_c {EK, finish fast write}. 
   The segment is not valid until ekffld_c has been called. 
   The EK system supports only one fast write at a time.  It is 
   not possible use the fast write routines to simultaneously write 
   multiple segments, either in the same EK file or in different 


   1)  Suppose we have an E-kernel named ORDER_DB.EK which contains 
       records of orders for data products.  The E-kernel has a 
       table called DATAORDERS that consists of the set of columns 
       listed below: 
             Column Name     Data Type 
             -----------     --------- 
             ORDER_ID        INTEGER 
             CUSTOMER_ID     INTEGER 
             LAST_NAME       CHARACTER*(*) 
             FIRST_NAME      CHARACTER*(*) 
             ORDER_DATE      TIME 
             COST            DOUBLE PRECISION 
       The order database also has a table of items that have been 
       ordered.  The columns of this table are shown below: 
             Column Name     Data Type 
             -----------     --------- 
             ITEM_ID         INTEGER 
             ORDER_ID        INTEGER 
             ITEM_NAME       CHARACTER*(*) 
             DESCRIPTION     CHARACTER*(*) 
             PRICE           DOUBLE PRECISION 
       We'll suppose that the file ORDER_DB.EK contains two segments, 
       the first containing the DATAORDERS table and the second 
       containing the DATAITEMS table. 
       Below, we show how we'd open a new EK file and create the 
       first of the segments described above. 

       #include "SpiceUsr.h"
       #include <stdio.h>
       void main()
          #define  CNMLEN      ( CSPICE_EK_COL_NAM_LEN + 1 )
          #define  DECLEN        201
          #define  EKNAME        "order_db.ek"
          #define  FNMLEN        50
          #define  IFNAME        "Test EK/Created 20-SEP-1995"
          #define  LNMLEN        50
          #define  LSK           "leapseconds.ker"
          #define  NCOLS         6
          #define  NRESVC        0
          #define  NROWS         9
          #define  TABLE         "DATAORDERS"
          #define  TNMLEN        CSPICE_EK_TAB_NAM_LEN
          #define  UTCLEN        30
          Local variables
          SpiceBoolean            nlflgs [ NROWS  ];
          SpiceChar               cdecls  [ NCOLS ] [ DECLEN ];
          SpiceChar               cnames  [ NCOLS ] [ CNMLEN ];
          SpiceChar               fnames  [ NROWS ] [ FNMLEN ];
          SpiceChar               lnames  [ NROWS ] [ LNMLEN ];
          SpiceChar               dateStr [ UTCLEN ];
          SpiceDouble             costs  [ NROWS ];
          SpiceDouble             ets    [ NROWS ];
          SpiceInt                cstids [ NROWS ];
          SpiceInt                ordids [ NROWS ];
          SpiceInt                handle;
          SpiceInt                i;
          SpiceInt                rcptrs [ NROWS ];
          SpiceInt                segno;
          SpiceInt                sizes  [ NROWS ];
          SpiceInt                wkindx [ NROWS ];
          Load a leapseconds kernel for UTC/ET conversion.
          furnsh_c ( LSK );
          Open a new EK file.  For simplicity, we will not 
          reserve any space for the comment area, so the 
          number of reserved comment characters is zero. 
          The constant IFNAME is the internal file name. 
          ekopn_c ( EKNAME, IFNAME, NRESVC, &handle );
          Set up the table and column names and declarations 
          for the DATAORDERS segment.  We'll index all of 
          the columns.  All columns are scalar, so we omit 
          the size declaration.  Only the COST column may take 
          null values. 
          strcpy ( cnames[0], "ORDER_ID"                           );
          strcpy ( cdecls[0], "DATATYPE = INTEGER, INDEXED = TRUE" );
          strcpy ( cnames[1], "CUSTOMER_ID"                        );
          strcpy ( cdecls[1], "DATATYPE = INTEGER, INDEXED = TRUE" );
          strcpy ( cnames[2], "LAST_NAME"                          ); 
          strcpy ( cdecls[2], "DATATYPE = CHARACTER*(*),"
                              "INDEXED  = TRUE"                    );
          strcpy ( cnames[3], "FIRST_NAME"                         );
          strcpy ( cdecls[3], "DATATYPE = CHARACTER*(*),"   
                              "INDEXED  = TRUE"                    );
          strcpy ( cnames[4], "ORDER_DATE"                         );
          strcpy ( cdecls[4], "DATATYPE = TIME, INDEXED  = TRUE"   );
          strcpy ( cnames[5], "COST"                               );
          strcpy ( cdecls[5], "DATATYPE = DOUBLE PRECISION,"   
                              "INDEXED  = TRUE,"             
                              "NULLS_OK = TRUE"                    );
          Start the segment.  We presume the number of  rows 
          of data is known in advance. 
          ekifld_c ( handle,  TABLE,   NCOLS,  NROWS,   CNMLEN,  
                     cnames,  DECLEN,  cdecls, &segno,  rcptrs );
          At this point, arrays containing data for the 
          segment's columns may be filled in.  The names 
          of the data arrays are shown below. 
             Column           Data array 
             "ORDER_ID"       ordids 
             "CUSTOMER_ID"    cstids 
             "LAST_NAME"      lnames 
             "FIRST_NAME"     fnames 
             "ORDER_DATE"     odates 
             "COST"           costs 
          The null flags array indicates which entries are null. 
          It is ignored for columns that don't allow null 
          values.  In this case, only the COST column allows 
          Fill in data arrays and null flag arrays here.  This code
          section would normally be replaced by calls to user functions
          returning column values.
          for ( i = 0;  i < NROWS;  i++ )
             ordids[i]  =  i;
             cstids[i]  =  i*100;
             costs [i]  =  (SpiceDouble) 100*i;
             sprintf  ( fnames[i], "Order %d Customer first name", i );
             sprintf  ( lnames[i], "Order %d Customer last name",  i );
             sprintf  ( dateStr,   "1998 Mar %d",                  i );
             utc2et_c ( dateStr, ets+i );
             nlflgs[i]  =  SPICEFALSE;
          nlflgs[1] = SPICETRUE;
          The sizes array shown below is ignored for scalar 
          and fixed-size array columns, so we need not 
          initialize it.  For variable-size arrays, the 
          Ith element of the sizes array must contain the size 
          of the Ith column entry in the column being written. 
          Normally, the sizes array would be reset for each 
          variable-size column. 
          Add the columns of data to the segment.  All of the 
          data for each column is written in one shot. 
          ekacli_c ( handle,  segno,   "order_id",    ordids, 
                     sizes,   nlflgs,  rcptrs,        wkindx ); 
          ekacli_c ( handle,  segno,   "customer_id", cstids,  
                     sizes,   nlflgs,  rcptrs,        wkindx ); 
          ekaclc_c ( handle,  segno,   "last_name",   LNMLEN,
                     lnames,  sizes,   nlflgs,        rcptrs,  wkindx ); 
          ekaclc_c ( handle,  segno,   "first_name",  FNMLEN,
                     fnames,  sizes,   nlflgs,        rcptrs,  wkindx ); 
          ekacld_c ( handle,  segno,   "order_date",  ets,  
                     sizes,   nlflgs,  rcptrs,        wkindx );
          ekacld_c ( handle,  segno,   "cost",        costs,  
                     sizes,   nlflgs,  rcptrs,        wkindx ); 
          Complete the segment.  The rcptrs array is that 
          returned by ekifld_c. 
          ekffld_c ( handle, segno, rcptrs ); 
          At this point, the second segment could be 
          created by an analogous process.  In fact, the 
          second segment could be created at any time; it is 
          not necessary to populate the first segment with 
          data before starting the second segment. 
          The file must be closed by a call to ekcls_c. 
          ekcls_c ( handle ); 


   1)  Only one segment can be created at a time using the fast 
       write routines. 
   2)  No other EK operation may interrupt a fast write.  For 
       example, it is not valid to issue a query while a fast write 
       is in progress. 




   N.J. Bachman   (JPL) 


   -CSPICE Version 2.3.1, 14-AUG-2006   (EDW)

      Replace mention of ldpool_c with furnsh_c.

   -CSPICE Version 2.3.0, 12-JUL-2002 (NJB)

      Call to C2F_CreateStrArr_Sig replaced with call to C2F_MapStrArr.

   -CSPICE Version 2.2.0,  10-JAN-2002 (NJB)

      Const-qualified input arrays.  Added casts to type (void *) 
      to expressions passed to free(), in order to suppress compilation
      warnings under MS Visual C++/C.

      Documentation change:  instances of the phrase "fast load"
      were replaced with "fast write."

      Corrected parameter names giving maximum table and column name

   -CSPICE Version 2.1.0, 14-FEB-2000 (NJB)

       Calls to C2F_CreateStrArr replaced with calls to error-signaling 
       version of this routine:  C2F_CreateStrArr_Sig.
   -CSPICE Version 2.0.0, 07-JUL-1999 (NJB)

       Output segment number segno is now mapped to C range.
   -CSPICE Version 1.0.0, 08-MAR-1999 (NJB)

       Based on SPICELIB Version 1.0.0, 25-OCT-1995 (NJB)


   start new E-kernel segment for fast writing 
   start new EK segment for fast writing 
Wed Apr  5 17:54:33 2017