This adds an installation step for PHP code for the tokenizer. The
PHP code is split in two parts. The updateable code is found in
lib-php. The tokenizer installs an additional script in the
project directory which then includes the code from lib-php and
defines all settings that are static to the database. The website
code then always includes the PHP from the project directory.
These tables have never been actively maintained and the code is
completely untested. With the upcomming changes, it is unlikely
that the code remains usable.
This removes the aux tables and all code that references them.
Disabling query reversal is no longer possible in the configuration,
so there is no need to keep this as an option. Reversal is
automatically disabled for structured search only.
Usually we don't narrow down search results by house number when
only a street name is given because there may be a lot of rows
to cross check when the street name is very frequent. However,
when it is known to be rare, the housenumber check may be done
anyway.
Fixes#2238.
Always look up the closest housenumber before looking up
interpolations. This ensures that closer housenumbers are
preferred over interpolations.
Fixes#2214.
Adds a general higher penalty for special search term and an
additional one if the term is anywhere but the beginning or the
end. Also housenumbers and special searches together are less
likely.
When we are in the final iteration of the search groups, it is not
possible to further delay the results. Unconditionally use the
results with the best rank instead.
Full word terms are already preferred for the name part. Adding
only one-word partials to the address, makes it impossible to
give a similar preference for the address part. Each term adds
a rank penalty. The problem here is that we interpret the query
forwards and backwards. Having different penalty systems for
name and address means that the same term ends up with different
penalties and that often leads to interpretations of the wrong
direction being in the way.
Also switches to jinja-based preprocessing, which allows to
simplify the SQL files. Use 'if not exists' where possible
so that the step can be rerun to fix missing indexes.
