Hi, checked in some improvements for reversedb:
- improved speed (significantly for large schemas) of XML generation by writing directly to a PrintWriter instead of concatenating Strings - updated XML generation; id for fields and classes are no longer created, references are by name now - improved error reporting when connecting to database - minor improvements here and there I tried to reproduce some of the bug reports of the recent weeks, unfortunately I couldn't confirm any of them. I also do not have databases other than Oracle to test with right now, so I haven't looked into problems with other databases. The generated XML usually validates against the DTD, only setting the database type is usually required. If you have some special fields in your schema (e.g. BLOB or NVARCHAR2 in Oracle) the JDBC type reported by the JDBC driver sometimes is OTHER. In this case you have to manually update the field type. If your table or field names in the database contain "special" characters like $, this is not encoded properly therefore rejected by the parser. In this case you also have to manually update your XML. regards, Florian +------------------------------------------+ | citec.at | | Software Application Development | | Sonnleithnergasse 2/13 | | A-1100 Wien | | | | Email: [EMAIL PROTECTED] | | Fon: +43 (1) 920 45 49 | | Fax: +43 (1) 920 45 49 | | Mob: +43 (699) 102 53 901 | +------------------------------------------+ --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
