[pylint] searx_extra/update/update_osm_keys_tags.py

BTW: move some comments into script's  doc-string

Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
This commit is contained in:
Markus Heiser 2021-06-06 12:58:53 +02:00
parent 5cf1ae2672
commit 8871e39122
1 changed files with 53 additions and 48 deletions

View File

@ -1,12 +1,43 @@
#!/usr/bin/env python #!/usr/bin/env python
""" # lint: pylint
Fetch OSM keys and tags # pylint: disable=missing-function-docstring
"""Fetch OSM keys and tags.
to get the i18n names, the scripts uses query.wikidata.org To get the i18n names, the scripts uses `Wikidata Query Service`_ instead of for
instead of for example https://taginfo.openstreetmap.org/taginfo/apidoc example `OSM tags API`_ (sidenote: the actual change log from
map.atownsend.org.uk_ might be useful to normalize OSM tags)
.. _Wikidata Query Service: https://query.wikidata.org/
.. _OSM tags API: https://taginfo.openstreetmap.org/taginfo/apidoc
.. _map.atownsend.org.uk: https://map.atownsend.org.uk/maps/map/changelog.html
:py:obj:`SPARQL_TAGS_REQUEST` :
Wikidata SPARQL query that returns *type-categories* and *types*. The
returned tag is ``Tag:{category}={type}`` (see :py:func:`get_tags`).
Example:
- https://taginfo.openstreetmap.org/tags/building=house#overview
- https://wiki.openstreetmap.org/wiki/Tag:building%3Dhouse
at the bottom of the infobox (right side), there is a link to wikidata:
https://www.wikidata.org/wiki/Q3947
see property "OpenStreetMap tag or key" (P1282)
- https://wiki.openstreetmap.org/wiki/Tag%3Abuilding%3Dbungalow
https://www.wikidata.org/wiki/Q850107
:py:obj:`SPARQL_KEYS_REQUEST` :
Wikidata SPARQL query that returns *keys*. Example with "payment":
- https://wiki.openstreetmap.org/wiki/Key%3Apayment
at the bottom of infobox (right side), there is a link to wikidata:
https://www.wikidata.org/wiki/Q1148747
link made using the "OpenStreetMap tag or key" property (P1282)
to be confirm: there is a one wiki page per key ?
- https://taginfo.openstreetmap.org/keys/payment#values
- https://taginfo.openstreetmap.org/keys/payment:cash#values
``rdfs:label`` get all the labels without language selection
(as opposed to SERVICE ``wikibase:label``).
https://map.atownsend.org.uk/maps/map/changelog.html (the actual change log)
might be useful to normalize OSM tags
""" """
import json import json
@ -19,50 +50,28 @@ from searx.engines.wikidata import send_wikidata_query
from searx.languages import language_codes from searx.languages import language_codes
from searx.engines.openstreetmap import get_key_rank, VALUE_TO_LINK from searx.engines.openstreetmap import get_key_rank, VALUE_TO_LINK
SPARQL_TAGS_REQUEST = """
# nominatim return type category and type
# the tag is "Tag:{category}={type}"
# Example:
# * https://taginfo.openstreetmap.org/tags/building=house#overview
# * https://wiki.openstreetmap.org/wiki/Tag:building%3Dhouse
# at the bottom of the infobox (right side), there is a link to wikidata:
# https://www.wikidata.org/wiki/Q3947
# see property "OpenStreetMap tag or key" (P1282)
# * https://wiki.openstreetmap.org/wiki/Tag%3Abuilding%3Dbungalow
# https://www.wikidata.org/wiki/Q850107
SARQL_TAGS_REQUEST = """
SELECT ?tag ?item ?itemLabel WHERE { SELECT ?tag ?item ?itemLabel WHERE {
?item wdt:P1282 ?tag . ?item wdt:P1282 ?tag .
?item rdfs:label ?itemLabel . ?item rdfs:label ?itemLabel .
FILTER(STRSTARTS(?tag, 'Tag')) FILTER(STRSTARTS(?tag, 'Tag'))
} }
GROUP BY ?tag ?item ?itemLabel GROUP BY ?tag ?item ?itemLabel
ORDER BY ?tag ?item ?itemLabel ORDER BY ?tag ?item ?itemLabel
""" """
# keys SPARQL_KEYS_REQUEST = """
# Example with "payment"":
# * https://wiki.openstreetmap.org/wiki/Key%3Apayment
# at the bottom of infobox (right side), there is a link to wikidata:
# https://www.wikidata.org/wiki/Q1148747
# link made using the "OpenStreetMap tag or key" property (P1282)
# to be confirm: there is a one wiki page per key ?
# * https://taginfo.openstreetmap.org/keys/payment#values
# * https://taginfo.openstreetmap.org/keys/payment:cash#values
#
# rdfs:label get all the labels without language selection
# (as opposed to SERVICE wikibase:label)
SARQL_KEYS_REQUEST = """
SELECT ?key ?item ?itemLabel WHERE { SELECT ?key ?item ?itemLabel WHERE {
?item wdt:P1282 ?key . ?item wdt:P1282 ?key .
?item rdfs:label ?itemLabel . ?item rdfs:label ?itemLabel .
FILTER(STRSTARTS(?key, 'Key')) FILTER(STRSTARTS(?key, 'Key'))
} }
GROUP BY ?key ?item ?itemLabel GROUP BY ?key ?item ?itemLabel
ORDER BY ?key ?item ?itemLabel ORDER BY ?key ?item ?itemLabel
""" """
LANGUAGES = [l[0].lower() for l in language_codes] LANGUAGES = [l[0].lower() for l in language_codes]
PRESET_KEYS = { PRESET_KEYS = {
('wikidata',): {'en': 'Wikidata'}, ('wikidata',): {'en': 'Wikidata'},
('wikipedia',): {'en': 'Wikipedia'}, ('wikipedia',): {'en': 'Wikipedia'},
@ -71,11 +80,11 @@ PRESET_KEYS = {
('fax',): {'en': 'Fax'}, ('fax',): {'en': 'Fax'},
('internet_access', 'ssid'): {'en': 'Wi-Fi'}, ('internet_access', 'ssid'): {'en': 'Wi-Fi'},
} }
INCLUDED_KEYS = { INCLUDED_KEYS = {
('addr', ) ('addr', )
} }
def get_preset_keys(): def get_preset_keys():
results = collections.OrderedDict() results = collections.OrderedDict()
for keys, value in PRESET_KEYS.items(): for keys, value in PRESET_KEYS.items():
@ -85,10 +94,9 @@ def get_preset_keys():
r.setdefault('*', value) r.setdefault('*', value)
return results return results
def get_keys(): def get_keys():
results = get_preset_keys() results = get_preset_keys()
response = send_wikidata_query(SARQL_KEYS_REQUEST) response = send_wikidata_query(SPARQL_KEYS_REQUEST)
for key in response['results']['bindings']: for key in response['results']['bindings']:
keys = key['key']['value'].split(':')[1:] keys = key['key']['value'].split(':')[1:]
@ -136,7 +144,7 @@ def get_keys():
def get_tags(): def get_tags():
results = collections.OrderedDict() results = collections.OrderedDict()
response = send_wikidata_query(SARQL_TAGS_REQUEST) response = send_wikidata_query(SPARQL_TAGS_REQUEST)
for tag in response['results']['bindings']: for tag in response['results']['bindings']:
tag_names = tag['tag']['value'].split(':')[1].split('=') tag_names = tag['tag']['value'].split(':')[1].split('=')
if len(tag_names) == 2: if len(tag_names) == 2:
@ -149,7 +157,6 @@ def get_tags():
results.setdefault(tag_category, {}).setdefault(tag_type, {}).setdefault(lang, label) results.setdefault(tag_category, {}).setdefault(tag_type, {}).setdefault(lang, label)
return results return results
def optimize_data_lang(translations): def optimize_data_lang(translations):
language_to_delete = [] language_to_delete = []
# remove "zh-hk" entry if the value is the same as "zh" # remove "zh-hk" entry if the value is the same as "zh"
@ -174,14 +181,12 @@ def optimize_data_lang(translations):
for language in language_to_delete: for language in language_to_delete:
del translations[language] del translations[language]
def optimize_tags(data): def optimize_tags(data):
for v in data.values(): for v in data.values():
for translations in v.values(): for translations in v.values():
optimize_data_lang(translations) optimize_data_lang(translations)
return data return data
def optimize_keys(data): def optimize_keys(data):
for k, v in data.items(): for k, v in data.items():
if k == '*': if k == '*':
@ -190,15 +195,15 @@ def optimize_keys(data):
optimize_keys(v) optimize_keys(v)
return data return data
def get_osm_tags_filename(): def get_osm_tags_filename():
return Path(searx_dir) / "data" / "osm_keys_tags.json" return Path(searx_dir) / "data" / "osm_keys_tags.json"
if __name__ == '__main__':
set_timeout_for_thread(60) set_timeout_for_thread(60)
result = { result = {
'keys': optimize_keys(get_keys()), 'keys': optimize_keys(get_keys()),
'tags': optimize_tags(get_tags()), 'tags': optimize_tags(get_tags()),
} }
with open(get_osm_tags_filename(), 'w') as f: with open(get_osm_tags_filename(), 'w') as f:
json.dump(result, f, indent=4, ensure_ascii=False) json.dump(result, f, indent=4, ensure_ascii=False)