Create a more effective implementation of XClob::length() and
XClob::getSubString() methods, where string is read segment-by-segment
instead of reading the whole underlying blob. That way it is possible to
handle big texts which would not fit into memory.

Also allow reading Clob data from a resultset with getString() and
writing it in a prepared statement with setString().

Implement XPreparedStatement::setClob(). Also implement a private
version of setClob() for creating a clob from OUString:

Allow the creation of a clob column with GUI by adding a new type in
ODataBaseMetaData::getTypeInfo().

Change-Id: Ibcbbdd80e8eed5e2a3fe55b0fa196401f1bcbcdf
Reviewed-on: https://gerrit.libreoffice.org/47093Reviewed-by: Tamás Bunth <btomi96@gmail.com>
Tested-by: Tamás Bunth <btomi96@gmail.com>