fix grave schema editing bug

Might be related to #82.

The SchemaBuilder expects input columns indexed by their colref number,
but Schema->columns is indexed by sort. Thus each edit possibly juggled

fix grave schema editing bug

Might be related to #82.

The SchemaBuilder expects input columns indexed by their colref number,
but Schema->columns is indexed by sort. Thus each edit possibly juggled
field definitions around.

show more ...

fixed summarize option. closes #83

fixed grouping in search

refacored tests to use StructTest as base class

this makes schema setup and data loading much easier and gets rid of a
lot of duplicate code

catch JSON error in importer

the column array should start at 0 and be sorted correctly

For some reason we stored them in column ref order. This might break
things that are not covered by tests but I hope not.

removed distinct feature from search

this was added for tag autocompletion but does not work as expected.
should be implemented in tag type directly

check for page existance in search. fixes #77

This requires an uptodate sqlite plugin

fix error when an aggregation tries to select a nonexisting column

print stacktrace on allow debug

adjust cachung if an aggregation depends on dynamic values #75

fixed struct filter vars and added tests #75

JSON_PRETTY_PRINT is not available in PHP 5.3

added Schema Import #32

This also adds a nice interface for that to the schema editor

allow exposting struct schemas as JSON #32

support for dynamic filter place holders #75

tests for struct stuff still missing
no cache handling, yet

avoid errors when no schema or column is given

handle page moves #72

added hidden functionality #15

allow regexps for page assignments. #70

moved pseudo syntax generation to SchemaData

fixed SQL when only multi values are selected. related to #64

allow validation to modify the value before saving

correctly handle multi val columns. fixes #64

now group_concat is used

no special handling for multi vals on sorting. fixes #61

since we have a copy of the first value now, we can simply sort by the
single value column. This will only work for data saved from now on.