You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The normalization module is one of the parts of Nominatim that make installation and deployment very difficult. It is also outdated and missing a lot scripts. We should get rid of the normalization module and do it in the frontend code instead. There we can easily make use of libICU or similar and will be flexible to add customized normalization modules.
There are two parts in the code where normalization is applied:
Import: currently the normalization is done as part of the placex update trigger. To move the normalization out, we need to do the normalization in the indexing code of the nominatim.py module. So for each place we need to get the possible names out of the database, normalize them and reinsert them for further processing in the trigger code.
Query: the normalization is the first stage of forward query processing (see make_standard_name()) in lib/Geocode.php.
The tricky part here is that the two parts are written in two different languages. Luckily both have decent support for libICU. Still we either first switch to Python for everything or do the normalization twice for a while.
The text was updated successfully, but these errors were encountered:
The normalization module is one of the parts of Nominatim that make installation and deployment very difficult. It is also outdated and missing a lot scripts. We should get rid of the normalization module and do it in the frontend code instead. There we can easily make use of libICU or similar and will be flexible to add customized normalization modules.
There are two parts in the code where normalization is applied:
Import: currently the normalization is done as part of the placex update trigger. To move the normalization out, we need to do the normalization in the indexing code of the nominatim.py module. So for each place we need to get the possible names out of the database, normalize them and reinsert them for further processing in the trigger code.
Query: the normalization is the first stage of forward query processing (see
make_standard_name()
) in lib/Geocode.php.The tricky part here is that the two parts are written in two different languages. Luckily both have decent support for libICU. Still we either first switch to Python for everything or do the normalization twice for a while.
The text was updated successfully, but these errors were encountered: