Datacage: Introduced <dc:group epxr="xpath" type="type"> ... </dc:group> and XPath function dc:group-key(). This splits the current result set into groups formed by expr. The type defaults to string. Afterwards all these groups are iterated by there natural order. The dc:group-key() gives access to the result of the grouping expression that forms a group. Say, you have a result set like this: name | description -----+------------ a | foo a | bar b | baz b | bla c | blub you can use: <dc:group expr="$name"> <group name="{dc:group-key()}"> <dc:for-each> <description value="{$description}"/> </dc:for-each> </group> </dc:group> to create: <group name="a"> <description name="foo"/> <description name="bar"/> </group> <group name="b"> <description name="baz"/> <description name="bla"/> </group> <group name="c"> <description name="blub"/> </group>

#!/bin/bash

mkdir -p datacagedb

DIR=`dirname $0`
DIR=`readlink -f "$DIR"`

CLASSPATH=
for l in `find "$DIR" -name \*.jar -print`; do
   CLASSPATH=$CLASSPATH:$l
done

export CLASSPATH
if [ $# != 1 ]; then
    echo "Usage: $0 <schema_file>"
fi

java org.h2.tools.RunScript \
    -url jdbc:h2:`readlink -f datacagedb`/datacage \
    -script "$1"