for some reason, when i try to add an hsql driver, i get ClassNotFoundException, why? (i allready the jar).
You seem to have CSS turned off.
Please don't fill out this field.
what version of Hypersonic SQL do you use?
The driver name for v1.2 and earlier is "hSql.hDriver".
for v1.3 and later it is "org.hsql.jdbcDriver"
version 1.8..2 i tried using org.hsql.jdbcDriver
Make sure that the jar file you entered in the "JDBC driver JAR" field really exists. You will get a ClassNotFoundException if the file cannot be found.
it does , i simply copied the hsqldb.jar to the lib directory, and it works with oracle ojdbc14.jar
did you specified the file in the "JDBC driver JAR" field of the connection details form? If the file is in the tool's lib folder, you must enter 'lib/hsqldb.jar' there. Alternatively you could edit the jailerGUI.sh/bat file and add the driver jar to the classpath.
i tried either one, doesn't work.
What's the error message of the ClassNotFoundException? Is it the driver class 'org.hsql.jdbcDriver' which was not found?
The connection information is stored in a file ".connections" (starting with a dot). I you like, you can send it to email@example.com I will check it then.
thanks for the file.
It seems that the driver class name has been changed in v1.8.
Please try "org.hsqldb.jdbcDriver" instead of "org.hsql.jdbcDriver"
thanks it's working now, i suspect you' ll update it on your end.:)
let me explain my original intent, we have an application based on hibernate and oracle. the model is rather compelx, and usually self suffcient, (very easy to subset) but it's very difficult to create the data for it, as the data for it is being gathered from legacy systems. so i thought i would create a parrallal schema in hsql, and each time we have a defect related to a data, i would extract the data to an hsql data, attach the db file to the defect, this would also serve in unit testing (as this is one of the reasons we have none), and each time i'll need to use it, i'll simply re-load it into oracle using jailer again.
my question is this, can it be specfied the certian columns (usually primary keys) will not retain their original values, but recieve new ones?
can i use jailer in batch mode? (give a root parameter string value, and have it extract a subset data into hsql ).
I do not quite understand why you want to have a parallel hsql schema. Jailer creates SQL-DML scripts or DBUnit files, which also could be attached to defects and reloaded into your oracle DB.
It is possible to filter the exported data. There is a small description about column-filters at http://jailer.sourceforge.net/faq.html (4. FAQ)
Jailer offers a command-line interface with which it is possible to use it in batch mode. Call jailer.sh (or jailer.bat on windows platform) without arguments to get some help about the usage. With the next release 3.4.4, which will be released soon, there is even GUI-support for the assembling of the command-line arguments.
we cannot use db units as most of our work is done on server side (i.e our classes need to function under j2ee server, it's dumb framework thing). so i'm left with just data. generating dml would be cool, but i need something that will allow me to delete what ever sub set data that was there before, or at least create a new complete new one, without mixing new and old data. so let's say i'm starting to work on a defect i would like to have the ability to "clean" my bad data, and reload it again.
the SQL-DML willl probably contain the original PRIMARY key values, which will collide with the previous data.
another advandage for a seperate schema would be to just "look" at the data in question, with no ties to my db dev enviorment, or better yet ,create an application what will read the data and display it in a primitve way.
when i first run jailer on my schema it said that a few of my tables have no primary key, will that stop me? it was generated by hibernate, and i wouldn't want to stop messing it around.
btw are there an thoughts in the future about eclipse plug-in integration?
before you say, "i don't want to force eclipse on the users"(you'd be right) you don't need to. eclipse's RCP framework will allow you to minimize your install the configuration and turn into a regular java application not JDT or tooling stuff, but it will be boosted by it's EMF+GMF used under the DTP project and that would probably cut your code (regarding the building of the graph and other common db stuff) by 10 fold.
plus, becouse DTP is db agnostic you could use a model which actually SPANS across databases, and not be restricted to one connection.
how about jpa integration? , allowing jailer to read persistence.xml and annotated class to make the model?
trying to solve the current problem i tried to get updated source code for 3.4.3 version, but i couldn't find it , the branches seem to match old versions. (AND i saw a begining of google.code repository) .
you can always define a PK of a table in the tool's internal data model using the data model editor, if the database table doesn't have one. There is an entry in the FAQ list about this topic.
Eclipse integration might be an interesting option in the future, as well as jpa integration. It's already possible to use other meta-data sources than the JDBC driver's meta-data API by implementing the interface 'net.sf.jailer.modelbuilder.ModelElementFinder".
Each release zip file also contains the source code. The most current source code is available via the project's SVN repository, inside the 'trunk' folder.
Sign up for the SourceForge newsletter: