OpenEdge Development:<br />Progress Dynamics Administration Writing a dump program

This is a sample dump program using default code that you can use to export the contents of the gsm_menu_structure_item table:

gsmitout.p sample using the default code
/* gsmitout.p */ {db/icf/dfd/sitedataprocout.i &OutputFile = "'gsmitout.d'" &OutputTable = "gsm_menu_structure_item" &ObjField = "menu_structure_item_obj"}

This program assumes that the data to be dumped is from a table that has an object ID field of data type decimal (ObjField include file directive). The object ID’s mantissa is used to determine the site number of the data to be dumped. The OutputFile include file directive provides a character string name for the file to be dumped. All dumped data is written to the first directory named sitedatadump that can be found in the PROPATH. The OutputTable directive contains the name of the table to be exported.

Using the default sitedataprocout.i include file to provide all of the code is the easiest way to produce a dump program. This code automatically exports the specified data for all sites listed in the dumpconfig.txt file to the dump files that are later read by a load program.

If you must add any special customizations, you can use code based on the following example to dump the data:

gsmitout.p sample using customized code
/* Include the parameters required for this program / {db/icf/dfd/sitedatahdrout.i} / Find the record in the ttDumpFileLocation table for the output file / FIND ttDumpFileLocation WHERE ttDumpFileLocation.cDumpFile = {&OutputFile} NO-ERROR. / If no record exists, or the path is ?, we do not need to write this file out. Simply return to the caller. / IF NOT AVAILABLE(ttDumpFileLocation) OR ttDumpFileLocation.cDumpFilePath = ? THEN RETURN. / Open the stream that we will write the data out on and make sure we APPEND to it. The file may have been created on a previous call to this program. The caller will have ensured that the file was a new file before the first call / OUTPUT TO VALUE(ttDumpFileLocation.cDumpFilePath) APPEND. FOR EACH / table that we need to dump / WHERE (Table.ObjID - TRUNCATE(Table.ObjID,0)= pdSiteNo): / This where clause gets us only the records for the site number that was passed into this procedure as a parameter. */ EXPORT Table. END. OUTPUT CLOSE.

gsmitout.p sample using customized code

/* Include the parameters required for this program */ 
{db/icf/dfd/sitedatahdrout.i} 
/* Find the record in the ttDumpFileLocation table for the output file */ 
FIND ttDumpFileLocation  
  WHERE ttDumpFileLocation.cDumpFile = {&OutputFile} 
  NO-ERROR. 
/* If no record exists, or the path is ?, we do not need to write this file 
out. Simply return to the caller. */ 
IF NOT AVAILABLE(ttDumpFileLocation) OR 
   ttDumpFileLocation.cDumpFilePath = ? THEN 
  RETURN. 
/* Open the stream that we will write the data out on and make sure we APPEND 
to it. The file may have been created on a previous call to this program. The 
caller will have ensured that the file was a new file before the first call */ 
OUTPUT TO VALUE(ttDumpFileLocation.cDumpFilePath) APPEND. 
FOR EACH /* table that we need to dump */ 
  WHERE (Table.ObjID - TRUNCATE(Table.ObjID,0)= pdSiteNo): 
/* This where clause gets us only the records for the site number that was 
passed into this procedure as a parameter. */ 
  EXPORT Table. 
END. 
OUTPUT CLOSE.

The comments in the above code describe what must be in the code. Two parameters defined in sitedatahdrout.i are passed to this procedure when it is called. The first is the site number for which data should be dumped in the mantissa of the pdSiteNo decimal field. The second is the table containing the dump file locations. It might be that your dump program requires more than one export file. The dumpconfig.txt file provides for this by allowing you to refer to the export file by the name you specify. For more information, see the "Creating a dumpconfig.txt file" section.