move get_addressdata() implementation to Python

[nominatim.git] / nominatim / api / lookup.py

diff --git a/nominatim/api/lookup.py b/nominatim/api/lookup.py
index 27706ff306e0aaaaf7866e4233060401339cc407..e9181f473784aec219c91f08acc3708e7dd3e516 100644 (file)
--- a/nominatim/api/lookup.py
+++ b/nominatim/api/lookup.py
@@ -7,33 +7,25 @@
 """
 Implementation of place lookup by ID.
 """
 """
 Implementation of place lookup by ID.
 """

-from typing import Optional
+from typing import Optional, Callable, Tuple, Type
+import datetime as dt

 
 import sqlalchemy as sa
 
 
 import sqlalchemy as sa
 

-from nominatim.typing import SaColumn, SaLabel, SaRow
+from nominatim.typing import SaColumn, SaRow, SaSelect

 from nominatim.api.connection import SearchConnection
 import nominatim.api.types as ntyp
 import nominatim.api.results as nres
 from nominatim.api.logging import log
 
 from nominatim.api.connection import SearchConnection
 import nominatim.api.types as ntyp
 import nominatim.api.results as nres
 from nominatim.api.logging import log
 

-def _select_column_geometry(column: SaColumn,
-                            geometry_output: ntyp.GeometryFormat) -> SaLabel:
-    """ Create the appropriate column expression for selecting a
-        geometry for the details response.
-    """
-    if geometry_output & ntyp.GeometryFormat.GEOJSON:
-        return sa.literal_column(f"""
-                  ST_AsGeoJSON(CASE WHEN ST_NPoints({column.name}) > 5000
-                               THEN ST_SimplifyPreserveTopology({column.name}, 0.0001)
-                               ELSE {column.name} END)
-                  """).label('geometry_geojson')
+RowFunc = Callable[[Optional[SaRow], Type[nres.BaseResultT]], Optional[nres.BaseResultT]]
+
+GeomFunc = Callable[[SaSelect, SaColumn], SaSelect]

 
 

-    return sa.func.ST_GeometryType(column).label('geometry_type')

 
 
 async def find_in_placex(conn: SearchConnection, place: ntyp.PlaceRef,
 
 
 async def find_in_placex(conn: SearchConnection, place: ntyp.PlaceRef,

-                         details: ntyp.LookupDetails) -> Optional[SaRow]:
+                         add_geometries: GeomFunc) -> Optional[SaRow]:

     """ Search for the given place in the placex table and return the
         base information.
     """
     """ Search for the given place in the placex table and return the
         base information.
     """


@@ -46,9 +38,8 @@ async def find_in_placex(conn: SearchConnection, place: ntyp.PlaceRef,
                     t.c.importance, t.c.wikipedia, t.c.indexed_date,
                     t.c.parent_place_id, t.c.rank_address, t.c.rank_search,
                     t.c.linked_place_id,
                     t.c.importance, t.c.wikipedia, t.c.indexed_date,
                     t.c.parent_place_id, t.c.rank_address, t.c.rank_search,
                     t.c.linked_place_id,

-                    sa.func.ST_X(t.c.centroid).label('x'),
-                    sa.func.ST_Y(t.c.centroid).label('y'),
-                    _select_column_geometry(t.c.geometry, details.geometry_output))
+                    t.c.geometry.ST_Expand(0).label('bbox'),
+                    t.c.centroid)

 
     if isinstance(place, ntyp.PlaceID):
         sql = sql.where(t.c.place_id == place.place_id)
 
     if isinstance(place, ntyp.PlaceID):
         sql = sql.where(t.c.place_id == place.place_id)


@@ -63,11 +54,11 @@ async def find_in_placex(conn: SearchConnection, place: ntyp.PlaceRef,
     else:
         return None
 
     else:
         return None
 

-    return (await conn.execute(sql)).one_or_none()
+    return (await conn.execute(add_geometries(sql, t.c.geometry))).one_or_none()

 
 
 async def find_in_osmline(conn: SearchConnection, place: ntyp.PlaceRef,
 
 
 async def find_in_osmline(conn: SearchConnection, place: ntyp.PlaceRef,

-                          details: ntyp.LookupDetails) -> Optional[SaRow]:
+                          add_geometries: GeomFunc) -> Optional[SaRow]:

     """ Search for the given place in the osmline table and return the
         base information.
     """
     """ Search for the given place in the osmline table and return the
         base information.
     """


@@ -76,9 +67,7 @@ async def find_in_osmline(conn: SearchConnection, place: ntyp.PlaceRef,
     sql = sa.select(t.c.place_id, t.c.osm_id, t.c.parent_place_id,
                     t.c.indexed_date, t.c.startnumber, t.c.endnumber,
                     t.c.step, t.c.address, t.c.postcode, t.c.country_code,
     sql = sa.select(t.c.place_id, t.c.osm_id, t.c.parent_place_id,
                     t.c.indexed_date, t.c.startnumber, t.c.endnumber,
                     t.c.step, t.c.address, t.c.postcode, t.c.country_code,

-                    sa.func.ST_X(sa.func.ST_Centroid(t.c.linegeo)).label('x'),
-                    sa.func.ST_Y(sa.func.ST_Centroid(t.c.linegeo)).label('y'),
-                    _select_column_geometry(t.c.linegeo, details.geometry_output))
+                    t.c.linegeo.ST_Centroid().label('centroid'))

 
     if isinstance(place, ntyp.PlaceID):
         sql = sql.where(t.c.place_id == place.place_id)
 
     if isinstance(place, ntyp.PlaceID):
         sql = sql.where(t.c.place_id == place.place_id)


@@ -93,89 +82,157 @@ async def find_in_osmline(conn: SearchConnection, place: ntyp.PlaceRef,
     else:
         return None
 
     else:
         return None
 

-    return (await conn.execute(sql)).one_or_none()
+    return (await conn.execute(add_geometries(sql, t.c.linegeo))).one_or_none()

 
 
 async def find_in_tiger(conn: SearchConnection, place: ntyp.PlaceRef,
 
 
 async def find_in_tiger(conn: SearchConnection, place: ntyp.PlaceRef,

-                        details: ntyp.LookupDetails) -> Optional[SaRow]:
+                        add_geometries: GeomFunc) -> Optional[SaRow]:

     """ Search for the given place in the table of Tiger addresses and return
         the base information. Only lookup by place ID is supported.
     """
     """ Search for the given place in the table of Tiger addresses and return
         the base information. Only lookup by place ID is supported.
     """

+    if not isinstance(place, ntyp.PlaceID):
+        return None
+

     log().section("Find in TIGER table")
     t = conn.t.tiger
     log().section("Find in TIGER table")
     t = conn.t.tiger

+    parent = conn.t.placex

     sql = sa.select(t.c.place_id, t.c.parent_place_id,
     sql = sa.select(t.c.place_id, t.c.parent_place_id,

+                    parent.c.osm_type, parent.c.osm_id,

                     t.c.startnumber, t.c.endnumber, t.c.step,
                     t.c.postcode,
                     t.c.startnumber, t.c.endnumber, t.c.step,
                     t.c.postcode,

-                    sa.func.ST_X(sa.func.ST_Centroid(t.c.linegeo)).label('x'),
-                    sa.func.ST_Y(sa.func.ST_Centroid(t.c.linegeo)).label('y'),
-                    _select_column_geometry(t.c.linegeo, details.geometry_output))
-
-    if isinstance(place, ntyp.PlaceID):
-        sql = sql.where(t.c.place_id == place.place_id)
-    else:
-        return None
+                    t.c.linegeo.ST_Centroid().label('centroid'))\
+            .where(t.c.place_id == place.place_id)\
+            .join(parent, t.c.parent_place_id == parent.c.place_id, isouter=True)

 
 

-    return (await conn.execute(sql)).one_or_none()
+    return (await conn.execute(add_geometries(sql, t.c.linegeo))).one_or_none()

 
 
 async def find_in_postcode(conn: SearchConnection, place: ntyp.PlaceRef,
 
 
 async def find_in_postcode(conn: SearchConnection, place: ntyp.PlaceRef,

-                           details: ntyp.LookupDetails) -> Optional[SaRow]:
+                           add_geometries: GeomFunc) -> Optional[SaRow]:

     """ Search for the given place in the postcode table and return the
         base information. Only lookup by place ID is supported.
     """
     """ Search for the given place in the postcode table and return the
         base information. Only lookup by place ID is supported.
     """

+    if not isinstance(place, ntyp.PlaceID):
+        return None
+

     log().section("Find in postcode table")
     t = conn.t.postcode
     sql = sa.select(t.c.place_id, t.c.parent_place_id,
                     t.c.rank_search, t.c.rank_address,
                     t.c.indexed_date, t.c.postcode, t.c.country_code,
     log().section("Find in postcode table")
     t = conn.t.postcode
     sql = sa.select(t.c.place_id, t.c.parent_place_id,
                     t.c.rank_search, t.c.rank_address,
                     t.c.indexed_date, t.c.postcode, t.c.country_code,

-                    sa.func.ST_X(t.c.geometry).label('x'),
-                    sa.func.ST_Y(t.c.geometry).label('y'),
-                    _select_column_geometry(t.c.geometry, details.geometry_output))
+                    t.c.geometry.label('centroid')) \
+            .where(t.c.place_id == place.place_id)

 
 

-    if isinstance(place, ntyp.PlaceID):
-        sql = sql.where(t.c.place_id == place.place_id)
-    else:
-        return None
+    return (await conn.execute(add_geometries(sql, t.c.geometry))).one_or_none()
+
+
+async def find_in_all_tables(conn: SearchConnection, place: ntyp.PlaceRef,
+                             add_geometries: GeomFunc
+                            ) -> Tuple[Optional[SaRow], RowFunc[nres.BaseResultT]]:
+    """ Search for the given place in all data tables
+        and return the base information.
+    """
+    row = await find_in_placex(conn, place, add_geometries)
+    log().var_dump('Result (placex)', row)
+    if row is not None:
+        return row, nres.create_from_placex_row
+
+    row = await find_in_osmline(conn, place, add_geometries)
+    log().var_dump('Result (osmline)', row)
+    if row is not None:
+        return row, nres.create_from_osmline_row

 
 

-    return (await conn.execute(sql)).one_or_none()
+    row = await find_in_postcode(conn, place, add_geometries)
+    log().var_dump('Result (postcode)', row)
+    if row is not None:
+        return row, nres.create_from_postcode_row
+
+    row = await find_in_tiger(conn, place, add_geometries)
+    log().var_dump('Result (tiger)', row)
+    return row, nres.create_from_tiger_row

 
 
 
 

-async def get_place_by_id(conn: SearchConnection, place: ntyp.PlaceRef,
-                          details: ntyp.LookupDetails) -> Optional[nres.SearchResult]:
+async def get_detailed_place(conn: SearchConnection, place: ntyp.PlaceRef,
+                             details: ntyp.LookupDetails) -> Optional[nres.DetailedResult]:

     """ Retrieve a place with additional details from the database.
     """
     """ Retrieve a place with additional details from the database.
     """

-    log().function('get_place_by_id', place=place, details=details)
+    log().function('get_detailed_place', place=place, details=details)

 
     if details.geometry_output and details.geometry_output != ntyp.GeometryFormat.GEOJSON:
         raise ValueError("lookup only supports geojosn polygon output.")
 
 
     if details.geometry_output and details.geometry_output != ntyp.GeometryFormat.GEOJSON:
         raise ValueError("lookup only supports geojosn polygon output.")
 

-    row = await find_in_placex(conn, place, details)
-    if row is not None:
-        result = nres.create_from_placex_row(row)
-        log().var_dump('Result', result)
-        await nres.add_result_details(conn, result, details)
-        return result
+    if details.geometry_output & ntyp.GeometryFormat.GEOJSON:
+        def _add_geometry(sql: SaSelect, column: SaColumn) -> SaSelect:
+            return sql.add_columns(sa.literal_column(f"""
+                      ST_AsGeoJSON(CASE WHEN ST_NPoints({column.name}) > 5000
+                                   THEN ST_SimplifyPreserveTopology({column.name}, 0.0001)
+                                   ELSE {column.name} END)
+                       """).label('geometry_geojson'))
+    else:
+        def _add_geometry(sql: SaSelect, column: SaColumn) -> SaSelect:
+            return sql.add_columns(sa.func.ST_GeometryType(column).label('geometry_type'))

 
 

-    row = await find_in_osmline(conn, place, details)
-    if row is not None:
-        result = nres.create_from_osmline_row(row)
-        log().var_dump('Result', result)
-        await nres.add_result_details(conn, result, details)
-        return result
+    row_func: RowFunc[nres.DetailedResult]
+    row, row_func = await find_in_all_tables(conn, place, _add_geometry)

 
 

-    row = await find_in_postcode(conn, place, details)
-    if row is not None:
-        result = nres.create_from_postcode_row(row)
-        log().var_dump('Result', result)
-        await nres.add_result_details(conn, result, details)
-        return result
+    if row is None:
+        return None

 
 

-    row = await find_in_tiger(conn, place, details)
-    if row is not None:
-        result = nres.create_from_tiger_row(row)
-        log().var_dump('Result', result)
-        await nres.add_result_details(conn, result, details)
-        return result
+    result = row_func(row, nres.DetailedResult)
+    assert result is not None
+
+    # add missing details
+    assert result is not None
+    indexed_date = getattr(row, 'indexed_date', None)
+    if indexed_date is not None:
+        result.indexed_date = indexed_date.replace(tzinfo=dt.timezone.utc)
+
+    await nres.add_result_details(conn, [result], details)
+
+    return result
+
+
+async def get_simple_place(conn: SearchConnection, place: ntyp.PlaceRef,
+                           details: ntyp.LookupDetails) -> Optional[nres.SearchResult]:
+    """ Retrieve a place as a simple search result from the database.
+    """
+    log().function('get_simple_place', place=place, details=details)
+
+    def _add_geometry(sql: SaSelect, col: SaColumn) -> SaSelect:
+        if not details.geometry_output:
+            return sql
+
+        out = []
+
+        if details.geometry_simplification > 0.0:
+            col = sa.func.ST_SimplifyPreserveTopology(col, details.geometry_simplification)
+
+        if details.geometry_output & ntyp.GeometryFormat.GEOJSON:
+            out.append(sa.func.ST_AsGeoJSON(col).label('geometry_geojson'))
+        if details.geometry_output & ntyp.GeometryFormat.TEXT:
+            out.append(sa.func.ST_AsText(col).label('geometry_text'))
+        if details.geometry_output & ntyp.GeometryFormat.KML:
+            out.append(sa.func.ST_AsKML(col).label('geometry_kml'))
+        if details.geometry_output & ntyp.GeometryFormat.SVG:
+            out.append(sa.func.ST_AsSVG(col).label('geometry_svg'))
+
+        return sql.add_columns(*out)
+
+
+    row_func: RowFunc[nres.SearchResult]
+    row, row_func = await find_in_all_tables(conn, place, _add_geometry)
+
+    if row is None:
+        return None
+
+    result = row_func(row, nres.SearchResult)
+    assert result is not None
+
+    # add missing details
+    assert result is not None
+    if hasattr(row, 'bbox'):
+        result.bbox = ntyp.Bbox.from_wkb(row.bbox)
+
+    await nres.add_result_details(conn, [result], details)

 
 

-    # Nothing found under this ID.
-    return None
+    return result