Use search_name table for TIGER data imports on 'dropped' databases

[lonvia]

When somebody has run the import with --drop or dropped the update information at a later point, it is no longer possible to import TIGER data because it relies on the location_road_2 table. We could be using the search_nametable instead in that case to look up the street. It will be quite a bit slower (maybe a temporary special index would help?) but at least it would work.

[ChristophTito]

Hello, I am new to this project and would like to participate! As this is declared as a good first issue, I would like to try out if I can contribute something. Can you maybe provide some hints where to look first and some guidance how to tackle this problem? Best regards, Christoph

[lonvia]

This concerns the implementation of the nominatim add-data --tiger-data. The basic implementation can be found innominatim/tools/tiger_data.py. However, to solve this issue, you have to dive a bit into the SQL functions. The main part of inserting the tiger data is done in the tiger_line_import function. Starting in line 73 there are three functions used getNearestNamedRoadPlaceId, getNearestParallelRoadFeature and getNearestRoadPlaceId. These three functions determine the closest road that should be used as the parent road. These function can be replaced with equivalent queries of search_name (for getNearestNamedRoadPlaceId) and placex (for the other two functions which need a geometry).

[pratham-saraf]

Hello As i am new here i thought i could start by contributing to good first issues. As per the information provided by lonvia the link to the tiger_line_import function is not functional so can it be re-provided so i can look at the guidance material provided Thanks regards Pratham Saraf