Storing information in the database

Storing single ZEBRA datastructures

CALL CDSTOR (PATH,LADDR,LKYBK*,IUDIV,KEYS,CHOPT,IRC*)

PATH

Character string describing the pathname

LADDR

Address of the datastructure to be stored N.B. LADDR must be protected against relocations or garbage collection, e.g. by using a link area.

LKYBK

Returned address of the key bank (option C)

IUDIV

Division where the datastructure resides

KEYS

Vector of keys. This vector must include space for the standard system and experiment keys and validity range pairs. The user must fill in the validity range pairs and user keys, if any.

CHOPT

Character string with any of the following characters

' '

Datastructure is stored asis, i.e. uncompressed

'C'

create Node/Key data structure ala CDUSE

'D'

store only differences from existing object in directory specified by PATH

'F'

With option D, differences are calculated from an object with which all user keys match (FULL match)

'K'

Store only the keys (LADDR ignored)

'H'

Insertion time in input KEYS vector is to be honoured

'P'

Data is compressed (bit packing)

'Y'

Store complete datastructure. N.B. by default, only the bank at LADDR will be stored.

'Z'

Store only nonzero elements. An element is considered to be zero if its absolute value is less than DELTA (a directory constant set by CDMDIR on page .

IRC

Integer return code  

0

Normal completion

53

Path name does not exist in database

61

Too many keys

63

Database structure in memory clobbered

64

Error in MZCOPY while copying Data bank

Routine CDSTOR stores a ZEBRA datastructure in the database. The data, if packed, is stored with a precision determined by the directory constant IPREC as specified in the call to CDMDIR on page

If the pathname begins with a % character then it is assumed to be an alias and is automatically translated by the internal HEPDB routines.

Storing multiple ZEBRA datastructures

CALL CDSTOM (PATH,LADDR,LKYBK*,IUDIV,NWKEY,NOBJ,KEYO,KEYN,CHOPT,IRC*)

PATH

Character string describing the pathname

LADDR

Vector of NOBJ bank addresses, created by CDBOOK N.B. LADDR must be protected against relocations or garbage collection, e.g. by using a link area.

LKYBK

Output address of first key bank (if option C is specified)

IUDIV

Division index of the user data bank

NWKEY

Number of keys associated with the data banks

NOBJ

Number of objects to be stored. The key vectors KEYO and KEYN must be dimensioned (NWKEY,NOBJ)

KEYO

Vector/matrix of old keys

KEYN

Vector/matrix of new keys

CHOPT

Character string with any of the following characters

' '

Datastructure(s) are stored asis, i.e. uncompressed

'C'

create Node/Key data structure ala CDUSE

'D'

store only differences from existing object in directory specified by PATH

'F'

With option D, differences are calculated from an object with which all user keys match (FULL match)

'K'

Store only the keys (LADDR ignored)

'H'

Insertion time in input keys vector is to be honoured. The insertion time is stored in KEYSN(KOFINS), packed using the routine CDPKTM.

'P'

Data is compressed (bit packing) according to the precision IPREC specified in the call to CDMDIR.

'Y'

Store complete datastructure. N.B. by default, only the bank at LADDR will be stored.

'Z'

Store only nonzero elements. An element is considered to be zero if its absolute value is less than DELTA (a directory constant set by CDMDIR on page

'R'

Replace existing objects as specified by the vector KEYO

IRC

Integer return code  

0

Normal completion

53

Path name does not exist in database

61

Too many keys

62

Too many keys with option N

63

Database structure in memory clobbered

64

Error in MZCOPY while copying Data bank

This routine permits multiple datastructures to be stored in a single call, or to replace one or more existing datastructures.

If the pathname begins with a % character then it is assumed to be an alias and is automatically translated by the internal HEPDB routines.