]> git.openstreetmap.org Git - nominatim.git/blobdiff - src/nominatim_db/tokenizer/icu_tokenizer.py
remove code only needed for older PostgreSQL/PostGIS versions
[nominatim.git] / src / nominatim_db / tokenizer / icu_tokenizer.py
index 7cd96d591fcec97483a721542829f4da90e52430..5595fcb2c604ae309bdb08dccf82e150764308dc 100644 (file)
@@ -13,12 +13,11 @@ from typing import Optional, Sequence, List, Tuple, Mapping, Any, cast, \
 import itertools
 import logging
 from pathlib import Path
-from textwrap import dedent
 
 from psycopg.types.json import Jsonb
 from psycopg import sql as pysql
 
-from ..db.connection import connect, Connection, Cursor, server_version_tuple,\
+from ..db.connection import connect, Connection, Cursor, \
                             drop_tables, table_exists, execute_scalar
 from ..config import Configuration
 from ..db.sql_preprocessor import SQLPreprocessor
@@ -33,10 +32,11 @@ DBCFG_TERM_NORMALIZATION = "tokenizer_term_normalization"
 
 LOG = logging.getLogger()
 
-WORD_TYPES =(('country_names', 'C'),
-             ('postcodes', 'P'),
-             ('full_word', 'W'),
-             ('housenumbers', 'H'))
+WORD_TYPES = (('country_names', 'C'),
+              ('postcodes', 'P'),
+              ('full_word', 'W'),
+              ('housenumbers', 'H'))
+
 
 def create(dsn: str, data_dir: Path) -> 'ICUTokenizer':
     """ Create a new instance of the tokenizer provided by this module.
@@ -55,7 +55,6 @@ class ICUTokenizer(AbstractTokenizer):
         self.data_dir = data_dir
         self.loader: Optional[ICURuleLoader] = None
 
-
     def init_new_db(self, config: Configuration, init_db: bool = True) -> None:
         """ Set up a new tokenizer for the database.
 
@@ -64,7 +63,6 @@ class ICUTokenizer(AbstractTokenizer):
         """
         self.loader = ICURuleLoader(config)
 
-        self._install_php(config.lib_dir.php, overwrite=True)
         self._save_config()
 
         if init_db:
@@ -72,7 +70,6 @@ class ICUTokenizer(AbstractTokenizer):
             self._setup_db_tables(config)
             self._create_base_indices(config, 'word')
 
-
     def init_from_project(self, config: Configuration) -> None:
         """ Initialise the tokenizer from the project directory.
         """
@@ -81,16 +78,12 @@ class ICUTokenizer(AbstractTokenizer):
         with connect(self.dsn) as conn:
             self.loader.load_config_from_db(conn)
 
-        self._install_php(config.lib_dir.php, overwrite=False)
-
-
     def finalize_import(self, config: Configuration) -> None:
         """ Do any required postprocessing to make the tokenizer data ready
             for use.
         """
         self._create_lookup_indices(config, 'word')
 
-
     def update_sql_functions(self, config: Configuration) -> None:
         """ Reimport the SQL functions for this tokenizer.
         """
@@ -98,14 +91,12 @@ class ICUTokenizer(AbstractTokenizer):
             sqlp = SQLPreprocessor(conn, config)
             sqlp.run_sql_file(conn, 'tokenizer/icu_tokenizer.sql')
 
-
     def check_database(self, config: Configuration) -> None:
         """ Check that the tokenizer is set up correctly.
         """
         # Will throw an error if there is an issue.
         self.init_from_project(config)
 
-
     def update_statistics(self, config: Configuration, threads: int = 2) -> None:
         """ Recompute frequencies for all name words.
         """
@@ -119,79 +110,37 @@ class ICUTokenizer(AbstractTokenizer):
                     cur.execute(pysql.SQL('SET max_parallel_workers_per_gather TO {}')
                                      .format(pysql.Literal(min(threads, 6),)))
 
-                if server_version_tuple(conn) < (12, 0):
-                    LOG.info('Computing word frequencies')
-                    drop_tables(conn, 'word_frequencies', 'addressword_frequencies')
-                    cur.execute("""CREATE TEMP TABLE word_frequencies AS
-                                     SELECT unnest(name_vector) as id, count(*)
-                                     FROM search_name GROUP BY id""")
-                    cur.execute('CREATE INDEX ON word_frequencies(id)')
-                    cur.execute("""CREATE TEMP TABLE addressword_frequencies AS
-                                     SELECT unnest(nameaddress_vector) as id, count(*)
-                                     FROM search_name GROUP BY id""")
-                    cur.execute('CREATE INDEX ON addressword_frequencies(id)')
-                    cur.execute("""CREATE OR REPLACE FUNCTION word_freq_update(wid INTEGER,
-                                                                               INOUT info JSONB)
-                                   AS $$
-                                   DECLARE rec RECORD;
-                                   BEGIN
-                                   IF info is null THEN
-                                     info = '{}'::jsonb;
-                                   END IF;
-                                   FOR rec IN SELECT count FROM word_frequencies WHERE id = wid
-                                   LOOP
-                                     info = info || jsonb_build_object('count', rec.count);
-                                   END LOOP;
-                                   FOR rec IN SELECT count FROM addressword_frequencies WHERE id = wid
-                                   LOOP
-                                     info = info || jsonb_build_object('addr_count', rec.count);
-                                   END LOOP;
-                                   IF info = '{}'::jsonb THEN
-                                     info = null;
-                                   END IF;
-                                   END;
-                                   $$ LANGUAGE plpgsql IMMUTABLE;
-                                """)
-                    LOG.info('Update word table with recomputed frequencies')
-                    drop_tables(conn, 'tmp_word')
-                    cur.execute("""CREATE TABLE tmp_word AS
-                                    SELECT word_id, word_token, type, word,
-                                           word_freq_update(word_id, info) as info
-                                    FROM word
-                                """)
-                    drop_tables(conn, 'word_frequencies', 'addressword_frequencies')
-                else:
-                    LOG.info('Computing word frequencies')
-                    drop_tables(conn, 'word_frequencies')
-                    cur.execute("""
-                      CREATE TEMP TABLE word_frequencies AS
-                      WITH word_freq AS MATERIALIZED (
-                               SELECT unnest(name_vector) as id, count(*)
-                                     FROM search_name GROUP BY id),
-                           addr_freq AS MATERIALIZED (
-                               SELECT unnest(nameaddress_vector) as id, count(*)
-                                     FROM search_name GROUP BY id)
-                      SELECT coalesce(a.id, w.id) as id,
-                             (CASE WHEN w.count is null THEN '{}'::JSONB
-                                  ELSE jsonb_build_object('count', w.count) END
-                              ||
-                              CASE WHEN a.count is null THEN '{}'::JSONB
-                                  ELSE jsonb_build_object('addr_count', a.count) END) as info
-                      FROM word_freq w FULL JOIN addr_freq a ON a.id = w.id;
-                      """)
-                    cur.execute('CREATE UNIQUE INDEX ON word_frequencies(id) INCLUDE(info)')
-                    cur.execute('ANALYSE word_frequencies')
-                    LOG.info('Update word table with recomputed frequencies')
-                    drop_tables(conn, 'tmp_word')
-                    cur.execute("""CREATE TABLE tmp_word AS
-                                    SELECT word_id, word_token, type, word,
-                                           (CASE WHEN wf.info is null THEN word.info
-                                            ELSE coalesce(word.info, '{}'::jsonb) || wf.info
-                                            END) as info
-                                    FROM word LEFT JOIN word_frequencies wf
-                                         ON word.word_id = wf.id
-                                """)
-                    drop_tables(conn, 'word_frequencies')
+                LOG.info('Computing word frequencies')
+                drop_tables(conn, 'word_frequencies')
+                cur.execute("""
+                  CREATE TEMP TABLE word_frequencies AS
+                  WITH word_freq AS MATERIALIZED (
+                           SELECT unnest(name_vector) as id, count(*)
+                                 FROM search_name GROUP BY id),
+                       addr_freq AS MATERIALIZED (
+                           SELECT unnest(nameaddress_vector) as id, count(*)
+                                 FROM search_name GROUP BY id)
+                  SELECT coalesce(a.id, w.id) as id,
+                         (CASE WHEN w.count is null THEN '{}'::JSONB
+                              ELSE jsonb_build_object('count', w.count) END
+                          ||
+                          CASE WHEN a.count is null THEN '{}'::JSONB
+                              ELSE jsonb_build_object('addr_count', a.count) END) as info
+                  FROM word_freq w FULL JOIN addr_freq a ON a.id = w.id;
+                  """)
+                cur.execute('CREATE UNIQUE INDEX ON word_frequencies(id) INCLUDE(info)')
+                cur.execute('ANALYSE word_frequencies')
+                LOG.info('Update word table with recomputed frequencies')
+                drop_tables(conn, 'tmp_word')
+                cur.execute("""CREATE TABLE tmp_word AS
+                                SELECT word_id, word_token, type, word,
+                                       (CASE WHEN wf.info is null THEN word.info
+                                        ELSE coalesce(word.info, '{}'::jsonb) || wf.info
+                                        END) as info
+                                FROM word LEFT JOIN word_frequencies wf
+                                     ON word.word_id = wf.id
+                            """)
+                drop_tables(conn, 'word_frequencies')
 
             with conn.cursor() as cur:
                 cur.execute('SET max_parallel_workers_per_gather TO 0')
@@ -204,8 +153,6 @@ class ICUTokenizer(AbstractTokenizer):
         self._create_lookup_indices(config, 'tmp_word')
         self._move_temporary_word_table('tmp_word')
 
-
-
     def _cleanup_housenumbers(self) -> None:
         """ Remove unused house numbers.
         """
@@ -239,8 +186,6 @@ class ICUTokenizer(AbstractTokenizer):
                                 (list(candidates.values()), ))
                 conn.commit()
 
-
-
     def update_word_tokens(self) -> None:
         """ Remove unused tokens.
         """
@@ -248,7 +193,6 @@ class ICUTokenizer(AbstractTokenizer):
         self._cleanup_housenumbers()
         LOG.warning("Tokenizer house-keeping done.")
 
-
     def name_analyzer(self) -> 'ICUNameAnalyzer':
         """ Create a new analyzer for tokenizing names and queries
             using this tokinzer. Analyzers are context managers and should
@@ -268,7 +212,6 @@ class ICUTokenizer(AbstractTokenizer):
         return ICUNameAnalyzer(self.dsn, self.loader.make_sanitizer(),
                                self.loader.make_token_analysis())
 
-
     def most_frequent_words(self, conn: Connection, num: int) -> List[str]:
         """ Return a list of the `num` most frequent full words
             in the database.
@@ -280,23 +223,6 @@ class ICUTokenizer(AbstractTokenizer):
                              ORDER BY count DESC LIMIT %s""", (num,))
             return list(s[0].split('@')[0] for s in cur)
 
-
-    def _install_php(self, phpdir: Optional[Path], overwrite: bool = True) -> None:
-        """ Install the php script for the tokenizer.
-        """
-        if phpdir is not None:
-            assert self.loader is not None
-            php_file = self.data_dir / "tokenizer.php"
-
-            if not php_file.exists() or overwrite:
-                php_file.write_text(dedent(f"""\
-                    <?php
-                    @define('CONST_Max_Word_Frequency', 10000000);
-                    @define('CONST_Term_Normalization_Rules', "{self.loader.normalization_rules}");
-                    @define('CONST_Transliteration', "{self.loader.get_search_rules()}");
-                    require_once('{phpdir}/tokenizer/icu_tokenizer.php');"""), encoding='utf-8')
-
-
     def _save_config(self) -> None:
         """ Save the configuration that needs to remain stable for the given
             database as database properties.
@@ -305,7 +231,6 @@ class ICUTokenizer(AbstractTokenizer):
         with connect(self.dsn) as conn:
             self.loader.save_config_to_db(conn)
 
-
     def _setup_db_tables(self, config: Configuration) -> None:
         """ Set up the word table and fill it with pre-computed word
             frequencies.
@@ -329,7 +254,6 @@ class ICUTokenizer(AbstractTokenizer):
             """)
             conn.commit()
 
-
     def _create_base_indices(self, config: Configuration, table_name: str) -> None:
         """ Set up the word table and fill it with pre-computed word
             frequencies.
@@ -350,21 +274,21 @@ class ICUTokenizer(AbstractTokenizer):
                                 column_type=ctype)
             conn.commit()
 
-
     def _create_lookup_indices(self, config: Configuration, table_name: str) -> None:
         """ Create additional indexes used when running the API.
         """
         with connect(self.dsn) as conn:
             sqlp = SQLPreprocessor(conn, config)
             # Index required for details lookup.
-            sqlp.run_string(conn, """
+            sqlp.run_string(
+                conn,
+                """
                 CREATE INDEX IF NOT EXISTS idx_{{table_name}}_word_id
                   ON {{table_name}} USING BTREE (word_id) {{db.tablespace.search_index}}
-            """,
-            table_name=table_name)
+                """,
+                table_name=table_name)
             conn.commit()
 
-
     def _move_temporary_word_table(self, old: str) -> None:
         """ Rename all tables and indexes used by the tokenizer.
         """
@@ -381,8 +305,6 @@ class ICUTokenizer(AbstractTokenizer):
             conn.commit()
 
 
-
-
 class ICUNameAnalyzer(AbstractAnalyzer):
     """ The ICU analyzer uses the ICU library for splitting names.
 
@@ -399,7 +321,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
 
         self._cache = _TokenCache()
 
-
     def close(self) -> None:
         """ Free all resources used by the analyzer.
         """
@@ -407,20 +328,17 @@ class ICUNameAnalyzer(AbstractAnalyzer):
             self.conn.close()
             self.conn = None
 
-
     def _search_normalized(self, name: str) -> str:
         """ Return the search token transliteration of the given name.
         """
         return cast(str, self.token_analysis.search.transliterate(name)).strip()
 
-
     def _normalized(self, name: str) -> str:
         """ Return the normalized version of the given name with all
             non-relevant information removed.
         """
         return cast(str, self.token_analysis.normalizer.transliterate(name)).strip()
 
-
     def get_word_token_info(self, words: Sequence[str]) -> List[Tuple[str, str, int]]:
         """ Return token information for the given list of words.
             If a word starts with # it is assumed to be a full name
@@ -452,8 +370,7 @@ class ICUNameAnalyzer(AbstractAnalyzer):
             part_ids = {r[0]: r[1] for r in cur}
 
         return [(k, v, full_ids.get(v, None)) for k, v in full_tokens.items()] \
-               + [(k, v, part_ids.get(v, None)) for k, v in partial_tokens.items()]
-
+            + [(k, v, part_ids.get(v, None)) for k, v in partial_tokens.items()]
 
     def normalize_postcode(self, postcode: str) -> str:
         """ Convert the postcode to a standardized form.
@@ -463,7 +380,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
         """
         return postcode.strip().upper()
 
-
     def update_postcodes_from_db(self) -> None:
         """ Update postcode tokens in the word table from the location_postcode
             table.
@@ -536,9 +452,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
             with self.conn.cursor() as cur:
                 cur.executemany("""SELECT create_postcode_word(%s, %s)""", terms)
 
-
-
-
     def update_special_phrases(self, phrases: Iterable[Tuple[str, str, str, str]],
                                should_replace: bool) -> None:
         """ Replace the search index for special phrases with the new phrases.
@@ -568,7 +481,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
         LOG.info("Total phrases: %s. Added: %s. Deleted: %s",
                  len(norm_phrases), added, deleted)
 
-
     def _add_special_phrases(self, cursor: Cursor,
                              new_phrases: Set[Tuple[str, str, str, str]],
                              existing_phrases: Set[Tuple[str, str, str, str]]) -> int:
@@ -588,10 +500,9 @@ class ICUNameAnalyzer(AbstractAnalyzer):
 
         return added
 
-
     def _remove_special_phrases(self, cursor: Cursor,
-                             new_phrases: Set[Tuple[str, str, str, str]],
-                             existing_phrases: Set[Tuple[str, str, str, str]]) -> int:
+                                new_phrases: Set[Tuple[str, str, str, str]],
+                                existing_phrases: Set[Tuple[str, str, str, str]]) -> int:
         """ Remove all phrases from the database that are no longer in the
             new phrase list.
         """
@@ -607,7 +518,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
 
         return len(to_delete)
 
-
     def add_country_names(self, country_code: str, names: Mapping[str, str]) -> None:
         """ Add default names for the given country to the search index.
         """
@@ -619,7 +529,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
                                      self.sanitizer.process_names(info)[0],
                                      internal=True)
 
-
     def _add_country_full_names(self, country_code: str, names: Sequence[PlaceName],
                                 internal: bool = False) -> None:
         """ Add names for the given country from an already sanitized
@@ -671,7 +580,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
                           """
                 cur.execute(sql, (country_code, list(new_tokens)))
 
-
     def process_place(self, place: PlaceInfo) -> Mapping[str, Any]:
         """ Determine tokenizer information about the given place.
 
@@ -694,7 +602,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
 
         return token_info.to_dict()
 
-
     def _process_place_address(self, token_info: '_TokenInfo',
                                address: Sequence[PlaceName]) -> None:
         for item in address:
@@ -707,12 +614,11 @@ class ICUNameAnalyzer(AbstractAnalyzer):
             elif item.kind == 'place':
                 if not item.suffix:
                     token_info.add_place(itertools.chain(*self._compute_name_tokens([item])))
-            elif not item.kind.startswith('_') and not item.suffix and \
-                 item.kind not in ('country', 'full', 'inclusion'):
+            elif (not item.kind.startswith('_') and not item.suffix and
+                  item.kind not in ('country', 'full', 'inclusion')):
                 token_info.add_address_term(item.kind,
                                             itertools.chain(*self._compute_name_tokens([item])))
 
-
     def _compute_housenumber_token(self, hnr: PlaceName) -> Tuple[Optional[int], Optional[str]]:
         """ Normalize the housenumber and return the word token and the
             canonical form.
@@ -748,7 +654,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
 
         return result
 
-
     def _retrieve_full_tokens(self, name: str) -> List[int]:
         """ Get the full name token for the given name, if it exists.
             The name is only retrieved for the standard analyser.
@@ -769,7 +674,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
 
         return full
 
-
     def _compute_name_tokens(self, names: Sequence[PlaceName]) -> Tuple[Set[int], Set[int]]:
         """ Computes the full name and partial name tokens for the given
             dictionary of names.
@@ -807,7 +711,6 @@ class ICUNameAnalyzer(AbstractAnalyzer):
 
         return full_tokens, partial_tokens
 
-
     def _add_postcode(self, item: PlaceName) -> Optional[str]:
         """ Make sure the normalized postcode is present in the word table.
         """
@@ -855,11 +758,9 @@ class _TokenInfo:
         self.address_tokens: Dict[str, str] = {}
         self.postcode: Optional[str] = None
 
-
     def _mk_array(self, tokens: Iterable[Any]) -> str:
         return f"{{{','.join((str(s) for s in tokens))}}}"
 
-
     def to_dict(self) -> Dict[str, Any]:
         """ Return the token information in database importable format.
         """
@@ -886,13 +787,11 @@ class _TokenInfo:
 
         return out
 
-
     def set_names(self, fulls: Iterable[int], partials: Iterable[int]) -> None:
         """ Adds token information for the normalised names.
         """
         self.names = self._mk_array(itertools.chain(fulls, partials))
 
-
     def add_housenumber(self, token: Optional[int], hnr: Optional[str]) -> None:
         """ Extract housenumber information from a list of normalised
             housenumbers.
@@ -902,7 +801,6 @@ class _TokenInfo:
             self.housenumbers.add(hnr)
             self.housenumber_tokens.add(token)
 
-
     def add_street(self, tokens: Iterable[int]) -> None:
         """ Add addr:street match terms.
         """
@@ -910,13 +808,11 @@ class _TokenInfo:
             self.street_tokens = set()
         self.street_tokens.update(tokens)
 
-
     def add_place(self, tokens: Iterable[int]) -> None:
         """ Add addr:place search and match terms.
         """
         self.place_tokens.update(tokens)
 
-
     def add_address_term(self, key: str, partials: Iterable[int]) -> None:
         """ Add additional address terms.
         """