X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/4fc61d260f1dbefeafff6e5b374852147f27afe0..815e3cccf6840d935e5a708588700cca23b976e9:/nominatim/tools/country_info.py?ds=sidebyside diff --git a/nominatim/tools/country_info.py b/nominatim/tools/country_info.py index 9ed666ff..ed04c2d5 100644 --- a/nominatim/tools/country_info.py +++ b/nominatim/tools/country_info.py @@ -7,11 +7,35 @@ """ Functions for importing and managing static country information. """ -import json import psycopg2.extras from nominatim.db import utils as db_utils from nominatim.db.connection import connect +from nominatim.errors import UsageError + +def _flatten_name_list(names): + if names is None: + return {} + + if not isinstance(names, dict): + raise UsageError("Expected key-value list for names in country_settings.py") + + flat = {} + for prefix, remain in names.items(): + if isinstance(remain, str): + flat[prefix] = remain + elif not isinstance(remain, dict): + raise UsageError("Entries in names must be key-value lists.") + else: + for suffix, name in remain.items(): + if suffix == 'default': + flat[prefix] = name + else: + flat[f'{prefix}:{suffix}'] = name + + return flat + + class _CountryInfo: """ Caches country-specific properties from the configuration file. @@ -27,31 +51,31 @@ class _CountryInfo: """ if not self._info: self._info = config.load_sub_configuration('country_settings.yaml') - self._key_prefix = 'name' - # Convert languages into a list for simpler handling. for prop in self._info.values(): + # Convert languages into a list for simpler handling. if 'languages' not in prop: prop['languages'] = [] elif not isinstance(prop['languages'], list): prop['languages'] = [x.strip() for x in prop['languages'].split(',')] - if 'names' not in prop: - prop['names'][self._key_prefix] = {} + prop['names'] = _flatten_name_list(prop.get('names')) + def items(self): """ Return tuples of (country_code, property dict) as iterable. """ return self._info.items() - def key_prefix(self): - """ Return the prefix that will be attached to the keys of the country - names values when storing them in the database + def get(self, country_code): + """ Get country information for the country with the given country code. """ - return self._key_prefix + return self._info.get(country_code, {}) + _COUNTRY_INFO = _CountryInfo() + def setup_country_config(config): """ Load country properties from the configuration file. Needs to be called before using any other functions in this @@ -72,9 +96,6 @@ def setup_country_tables(dsn, sql_dir, ignore_partitions=False): """ db_utils.execute_file(dsn, sql_dir / 'country_osm_grid.sql.gz') - def add_prefix_to_keys(names, prefix): - return {prefix+':'+k: v for k, v in names.items()} - params = [] for ccode, props in _COUNTRY_INFO.items(): if ccode is not None and props is not None: @@ -82,12 +103,13 @@ def setup_country_tables(dsn, sql_dir, ignore_partitions=False): partition = 0 else: partition = props.get('partition') - lang = props['languages'][0] if len(props['languages']) == 1 else None - name = add_prefix_to_keys(props.get('names').get(_COUNTRY_INFO.key_prefix()), _COUNTRY_INFO.key_prefix()) - name = json.dumps(name, ensure_ascii=False, separators=(', ', '=>')) - params.append((ccode, name[1:-1], lang, partition)) + lang = props['languages'][0] if len( + props['languages']) == 1 else None + + params.append((ccode, props['names'], lang, partition)) with connect(dsn) as conn: with conn.cursor() as cur: + psycopg2.extras.register_hstore(cur) cur.execute( """ CREATE TABLE public.country_name ( country_code character varying(2), @@ -113,8 +135,8 @@ def create_country_names(conn, tokenizer, languages=None): languages = languages.split(',') def _include_key(key): - return key == _COUNTRY_INFO.key_prefix() or \ - (key.startswith(_COUNTRY_INFO.key_prefix()+':') and (not languages or key[len(_COUNTRY_INFO.key_prefix())+1:] in languages)) + return ':' not in key or not languages or \ + key[key.index(':') + 1:] in languages with conn.cursor() as cur: psycopg2.extras.register_hstore(cur) @@ -124,14 +146,10 @@ def create_country_names(conn, tokenizer, languages=None): with tokenizer.name_analyzer() as analyzer: for code, name in cur: names = {'countrycode': code} - if code == 'gb': - names['short_name'] = 'UK' - if code == 'us': - names['short_name'] = 'United States' # country names (only in languages as provided) if name: - names.update(((k, v) for k, v in name.items() if _include_key(k))) + names.update({k : v for k, v in name.items() if _include_key(k)}) analyzer.add_country_names(code, names)