Background: After half an hour or so of fiddling, and thanks to some answers here, I was able to coax Spring Roo into reverse engineering the booktown sample PostgreSQL database. Thing is, the database has so many tables that when I tried to run a generated web app, it stack traced. No problem. Given the size of the database that's not unreasonable behavior, IMO.
Thing is, I tried to cut down the number of tables to a reasonable size using the --includeTables (and, later, the --excludeTables) options of the database reverse engineer command. But it didn't work. No syntax error was flagged but regardless what I specified all the tables were included. The reference materials are quite clear that the operand must be a space-separated list of tables, enclosed in double quotes, such as "Authors Books", which is what I used with --includeTables. With --excludeTables I listed every table but these two.
It seems pretty clear to me that I'm dealing with yet another Spring Roo bug. But before I spend the costly but necessary time to create a clean and reproducible JIRA report (using a much smaller database than booktown ), I'd like to narrow the boundaries of the bug by asking:
Question: Has anyone gotten --includeTables or its dual, --excludeTables, to function as documented?
If so, would you please describe the circumstances so that I can perhaps see how they differed from my booktown experience. And, have you tried your code under the current release, 1.2.5?
(I've found several apparent regression errors; that is, features that worked under earlier releases that don't work under 1.2.5--such as embedded media, which failed beginning with release 1.2.1, for which I've filed a JIRA).
Thanks muchly.
Yes, I've gotten this to work, the syntax is working for me even with 1.2.5. One thing to try would be to delete the dbre.xml file, and start fresh. You can also use "*", as in "MYBASETABLENAME*" to catch more tables at once in the exclude-