move abbreviation computation into import phase

[nominatim.git] / nominatim / db / status.py

diff --git a/nominatim/db/status.py b/nominatim/db/status.py
index a0454771e0e0653d386d480050082fa047e0ca2c..71e587875cd4db6c633d98b44118d6bed3202112 100644 (file)
--- a/nominatim/db/status.py
+++ b/nominatim/db/status.py
@@ -1,13 +1,15 @@
 """
 """

-Access and helper functions for the status table.
+Access and helper functions for the status and status log table.

 """
 import datetime as dt
 import logging
 import re
 
 """
 import datetime as dt
 import logging
 import re
 

-from ..tools.exec_utils import get_url
+from nominatim.tools.exec_utils import get_url
+from nominatim.errors import UsageError

 
 LOG = logging.getLogger()
 
 LOG = logging.getLogger()

+ISODATE_FORMAT = '%Y-%m-%dT%H:%M:%S'

 
 def compute_database_date(conn):
     """ Determine the date of the database from the newest object in the
 
 def compute_database_date(conn):
     """ Determine the date of the database from the newest object in the


@@ -15,11 +17,14 @@ def compute_database_date(conn):
     """
     # First, find the node with the highest ID in the database
     with conn.cursor() as cur:
     """
     # First, find the node with the highest ID in the database
     with conn.cursor() as cur:

-        osmid = cur.scalar("SELECT max(osm_id) FROM place WHERE osm_type='N'")
+        if conn.table_exists('place'):
+            osmid = cur.scalar("SELECT max(osm_id) FROM place WHERE osm_type='N'")
+        else:
+            osmid = cur.scalar("SELECT max(osm_id) FROM placex WHERE osm_type='N'")

 
         if osmid is None:
             LOG.fatal("No data found in the database.")
 
         if osmid is None:
             LOG.fatal("No data found in the database.")

-            raise RuntimeError("No data found in the database.")
+            raise UsageError("No data found in the database.")

 
     LOG.info("Using node id %d for timestamp lookup", osmid)
     # Get the node from the API to find the timestamp when it was created.
 
     LOG.info("Using node id %d for timestamp lookup", osmid)
     # Get the node from the API to find the timestamp when it was created.


@@ -31,21 +36,27 @@ def compute_database_date(conn):
     if match is None:
         LOG.fatal("The node data downloaded from the API does not contain valid data.\n"
                   "URL used: %s", node_url)
     if match is None:
         LOG.fatal("The node data downloaded from the API does not contain valid data.\n"
                   "URL used: %s", node_url)

-        raise RuntimeError("Bad API data.")
+        raise UsageError("Bad API data.")

 
 

-    LOG.debug("Found timestamp %s", match[1])
+    LOG.debug("Found timestamp %s", match.group(1))

 
 

-    return dt.datetime.fromisoformat(match[1]).replace(tzinfo=dt.timezone.utc)
+    return dt.datetime.strptime(match.group(1), ISODATE_FORMAT).replace(tzinfo=dt.timezone.utc)

 
 
 def set_status(conn, date, seq=None, indexed=True):
 
 
 def set_status(conn, date, seq=None, indexed=True):

-    """ Replace the current status with the given status.
+    """ Replace the current status with the given status. If date is `None`
+        then only sequence and indexed will be updated as given. Otherwise
+        the whole status is replaced.

     """
     """

-    assert date.tzinfo == dt.timezone.utc
+    assert date is None or date.tzinfo == dt.timezone.utc

     with conn.cursor() as cur:
     with conn.cursor() as cur:

-        cur.execute("TRUNCATE TABLE import_status")
-        cur.execute("""INSERT INTO import_status (lastimportdate, sequence_id, indexed)
-                       VALUES (%s, %s, %s)""", (date, seq, indexed))
+        if date is None:
+            cur.execute("UPDATE import_status set sequence_id = %s, indexed = %s",
+                        (seq, indexed))
+        else:
+            cur.execute("TRUNCATE TABLE import_status")
+            cur.execute("""INSERT INTO import_status (lastimportdate, sequence_id, indexed)
+                           VALUES (%s, %s, %s)""", (date, seq, indexed))

 
     conn.commit()
 
 
     conn.commit()
 


@@ -61,3 +72,21 @@ def get_status(conn):
 
         row = cur.fetchone()
         return row['lastimportdate'], row['sequence_id'], row['indexed']
 
         row = cur.fetchone()
         return row['lastimportdate'], row['sequence_id'], row['indexed']

+
+
+def set_indexed(conn, state):
+    """ Set the indexed flag in the status table to the given state.
+    """
+    with conn.cursor() as cur:
+        cur.execute("UPDATE import_status SET indexed = %s", (state, ))
+    conn.commit()
+
+
+def log_status(conn, start, event, batchsize=None):
+    """ Write a new status line to the `import_osmosis_log` table.
+    """
+    with conn.cursor() as cur:
+        cur.execute("""INSERT INTO import_osmosis_log
+                       (batchend, batchseq, batchsize, starttime, endtime, event)
+                       SELECT lastimportdate, sequence_id, %s, %s, now(), %s FROM import_status""",
+                    (batchsize, start, event))