Difference between revisions of "XML files"
DrewJensen (Talk | contribs) m (→Before XML) |
DrewJensen (Talk | contribs) m (→Before XML) |
||
Line 23: | Line 23: | ||
buildTextTable( Conn, "Employees" ) | buildTextTable( Conn, "Employees" ) | ||
buildTextTable( Conn, "Customers" ) | buildTextTable( Conn, "Customers" ) | ||
+ | buildTextTable( Conn, "Routes" ) | ||
conn.dispose | conn.dispose | ||
Line 47: | Line 48: | ||
</pre> | </pre> | ||
− | Well, almost just fine. When the routine is done there are four updated files in the same directory as the admin.odb file: txt_categories.csv, txt_customers.csv, txt_products.csv and txt_employees.csv. | + | Well, almost just fine. When the routine is done there are four updated files in the same directory as the admin.odb file: txt_categories.csv, txt_customers.csv, txt_products.csv and txt_employees.csv, txt_routes.csv. ( one note on this - if the table being exported is empty, then that file will not be created. ) |
There is a subtle bug in this code...or in Base? If the file admin.odb is not open when this libraries main function is run it runs with out a problem, as many times as you care to run it. | There is a subtle bug in this code...or in Base? If the file admin.odb is not open when this libraries main function is run it runs with out a problem, as many times as you care to run it. |
Revision as of 06:49, 3 July 2006
The xml files will be simple, because if for no other reason I am not all that a proficint XMLer.
Before XML
Well, not wanting to wait on doing first things first - like creating xml data schemas...how about a throw back to the day. Meaning that for the moment I think we want to create some text files that will load our truck database. I have gone ahead and created an admin.odb file, with a basic structure the same as the driver.odb database. In this I will just add text tables that are identical to the tables to be exported to each driver.odb file. These would be Customers, Categories, Products, Routes and Employees.
With the HSQL engine exporting a table to a text file could not be easier. These couple of basic lines wll do this just fine.
Sub Main BasicLibraries.LoadLibrary("Drews") ' the library just supplies the getConnection routine exportTables() End Sub sub exportTables dim conn conn = getConnection( "admin" ) buildTextTable( Conn, "Categories" ) buildTextTable( Conn, "Products" ) buildTextTable( Conn, "Employees" ) buildTextTable( Conn, "Customers" ) buildTextTable( Conn, "Routes" ) conn.dispose end sub sub buildTextTable( aConn as variant, aTblName as variant ) dim aTbl as variant dim stmnt dim strTableName as string dim strCmd as string aTbl = aConn.Tables.GetByName( aTblName ) stmnt = aConn.createStatement strTableName = "txt_" & aTbl.Name if aConn.Tables.HasByName( strTableName ) then aConn.Tables.DropByname( strTableName ) end if stmnt.executeUpdate( "DROP TABLE " & strTableName & " IF EXISTS") stmnt.Escapeprocessing = False stmnt.execute( "SELECT * INTO TEXT """ & strTableName _ & """ FROM """ & aTbl.Name & """" ) end sub
Well, almost just fine. When the routine is done there are four updated files in the same directory as the admin.odb file: txt_categories.csv, txt_customers.csv, txt_products.csv and txt_employees.csv, txt_routes.csv. ( one note on this - if the table being exported is empty, then that file will not be created. )
There is a subtle bug in this code...or in Base? If the file admin.odb is not open when this libraries main function is run it runs with out a problem, as many times as you care to run it.
However - if admin.odb is open ( this is for an embedded database at least ) then the following happens. The first time you run this code it runs fine. You get the text tables and the text files all created. Now the second time you run it an error will occur - stating that the insert into routine failed, because the text table already exists. Here is how to fix it - after the routine is run you must refresh the table view in the Base main window. If you don't do this the line aConn.Tables.HasByName fails to find the file. I suppose filing an issue is in order, mean time I will look for the proper command to use in a dispatch statment to refresh the table list, before we start as a work around.
That done, we can create our export files just that easily.
Next the mirror of these - the import routines for the driver.odb file.