|
Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.stsci.edu/~sontag/spicedocs/cspice/ekgd_c.html
Дата изменения: Sat Dec 17 06:08:53 2005 Дата индексирования: Sun Apr 10 22:51:49 2016 Кодировка: |
Return an element of an entry in a column of double precision type in a specified row.
EK
Variable I/O Description -------- --- -------------------------------------------------- selidx I Index of parent column in SELECT clause. row I Row to fetch from. elment I Index of element, within column entry, to fetch. ddata O Double precision element of column entry. null O Flag indicating whether column entry was null. found O Flag indicating whether column was present in row.
selidx is the SELECT clause index of the column to fetch
from. The range of selidx is from 0 to one less than
the number of columns in the SELECT clause.
row is the output row containing the entry to fetch
from. The range of row is from 0 to one less than
the number of rows satisfying the previous query.
elment is the index of the element of the column entry
to fetch. The normal range of elment is from 0 to
one less than the size of the column's entry, but
elment is allowed to exceed the number of elements in
the column entry; if it does, found is returned
as SPICEFALSE. This allows the caller to read data
from the column entry in a loop without checking the
number of available elements first.
Null values in variable-sized columns are
considered to have size 1.
ddata is the requested element of the specified column
entry. If the entry is null, ddata is undefined.
null is a logical flag indicating whether the entry
belonging to the specified column in the specified
row is null.
found is a logical flag indicating whether the specified
element was found. If the element does not exist,
found is returned as SPICEFALSE.
None.
This routine allows retrieval of data from double precision columns. This routine returns one element at a time in order to save the caller from imposing a limit on the size of the column entries that can be handled.
1) Suppose the EK table TAB contains the following columns:
Column name Data Type Size
----------- --------- ----
DP_COL_1 DP 1
DP_COL_2 DP VARIABLE
DP_COL_3 DP 10
Suppose the query
query = "SELECT DP_COL_1 FROM TAB"
is issued to ekfind_c via the call
ekfind_c ( query, lenout, nmrows, error, errmsg );
To fetch and dump column values from the rows that satisfy the
query, the loop below could be used. Note that we don't check
the found flags returned by ekgd_c since we know that every
entry in column DP_COL_1 contains one element.
/.
Since DP_COL_1 was the first column selected,
the selection index selidx is set to 0.
The column is scalar, so the element index eltidx
is set to 0. The variable nmrows is the number of
matching rows returned by ekfind_c.
./
selidx = 0;
eltidx = 0;
for ( row = 0; row < nmrows; row++ )
{
printf ( "\nRow = %d\n\n", row );
/.
Fetch values from column DP_COL_1.
./
ekgd_c ( selidx, row, eltidx,
dval, &isnull, &found );
if ( isnull )
{
printf ( "%s\n", "<null>" );
}
else
{
printf ( "%f\n", dval );
}
}
2) Suppose the EK table TAB is as in example 1, and we issue
the query
query = "SELECT DP_COL_1, DP_COL_2, DP_COL_3 FROM TAB"
to ekfind_c via the call
ekfind_c ( query, lenout, &nmrows, &error, errmsg );
To fetch and dump column values from the rows that satisfy the
query, the loop below could be used. Note that we don't check
the found flags returned by ekgd_c since we know in advance how
many elements are contained in each column entry we fetch.
for ( row = 0; row < nmrows; row++ )
{
printf ( "\nRow = %d\n\n", row );
/.
Fetch values from column DP_COL_1. Since
DP_COL_1 was the first column selected, the
selection index selidx is set to 0.
./
selidx = 0;
eltidx = 0;
ekgd_c ( selidx, row, eltidx,
dvals[0], &isnull, &found )
printf ( "\nColumn = DP_COL_1\n\n" );
if ( isnull )
{
printf ( "%s\n", "<null>" );
}
else
{
printf ( "%f\n", dvals[0] );
}
/.
Fetch values from column DP_COL_2 in the current
row. Since DP_COL_2 contains variable-size array
entries, we call eknelt_c to determine how many
elements to fetch.
./
selidx = 1;
eknelt_c ( selidx, row, &nelt );
eltidx = 0;
isnull = SPICEFALSE;
while ( ( eltidx < nelt ) && ( !isnull ) )
{
ekgd_c ( selidx, row, eltidx,
dvals[eltidx], &isnull, &found );
eltidx++;
/.
If the column entry is null, we'll be kicked
out of this loop after the first iteration.
./
}
printf ( "\nColumn = DP_COL_2\n\n" );
if ( isnull )
{
printf ( "%s\n", "<null>" );
}
else
{
for ( i = 0; i < nelt; i++ )
{
printf ( "%f\n", dvals[i] );
}
}
/.
Fetch values from column DP_COL_3 in the current
row. We need not call eknelt_c since we know how
many elements are in each column entry.
./
selidx = 2;
eltidx = 0;
isnull = SPICEFALSE;
while ( ( eltidx < 10 ) && ( !isnull ) )
{
ekgd_c ( selidx, row, eltidx,
dvals[eltidx], &isnull, &found );
eltidx++;
}
printf ( "\nColumn = DP_COL_3\n\n" );
if ( isnull )
{
printf ( "%s\n", "<null>" );
}
else
{
for ( i = 0; i < 10; i++ )
{
printf ( "%f\n", dvals[i] );
}
}
}
3) See the Examples section of the query routine ekfind_c
for an example in which the names and data types of the
columns from which to fetch data are not known in advance.
None.
1) If the input argument elment is less than 0, found is returned as
SPICEFALSE, and the error SPICE(INVALIDINDEX) is signalled.
However, elment is allowed to be greater than or equal to
the number of elements in the specified column entry; this allows
the caller to read data from the column entry in a loop without
checking the number of available elements first. If elment is
greater than or equal to the number of available elements, found
is returned as SPICEFALSE.
2) If selidx is outside of the range established by the
last query passed to eksrch_, the error SPICE(INVALIDINDEX)
will be signalled.
3) If the input argument row is less than 0 or greater than or
equal to the number of rows matching the query, found is returned
as SPICEFALSE, and the error SPICE(INVALIDINDEX) is signalled.
4) If the specified column does not have double precision type, the
error SPICE(INVALIDTYPE) is signalled.
5) If this routine is called when no E-kernels have been loaded,
the error SPICE(NOLOADEDFILES) is signalled.
The EK "query and fetch" suite of functions reads binary `sequence
component' EK files. In order for a binary EK file to be
accessible to this routine, the file must be `loaded' via a call
to the function eklef_c.
Text format EK files cannot be used by this routine; they must
first be converted by binary format by the NAIF Toolkit utility
SPACIT.
N.J. Bachman (JPL)
None.
-CSPICE Version 1.1.1, 09-FEB-2003 (EDW)
Minor edit to correct typo in header.
-CSPICE Version 1.1.0, 09-JUL-1998 (NJB)
Bug fix: now uses local logical variable to capture the
error flag value returned by the underlying f2c'd routine.
-CSPICE Version 1.0.0, 27-MAR-1998
Based on SPICELIB Version 1.1.0, 07-JUL-1996 (NJB)
fetch element from double precision column entry