Merge pull request #90 from searxng/enh-osm-engine

[enh] improve openstreetmap engine

Makefile

@@ -73,7 +73,7 @@ test.shell:
 
 MANAGE += buildenv
 MANAGE += babel.compile
-MANAGE += data.all data.languages data.useragents
+MANAGE += data.all data.languages data.useragents data.osm_keys_tags
 MANAGE += docs.html docs.live docs.gh-pages docs.prebuild docs.clean
 MANAGE += docker.build docker.push docker.buildx
 MANAGE += gecko.driver

manage

@@ -115,6 +115,7 @@ babel.compile() {
 data.all() {
     data.languages
     data.useragents
+    data.osm_keys_tags
     build_msg DATA "update searx/data/ahmia_blacklist.txt"
     pyenv.cmd python searx_extra/update/update_ahmia_blacklist.py
     build_msg DATA "update searx/data/wikidata_units.json"
@@ -140,6 +141,12 @@ data.useragents() {
     dump_return $?
 }
 
+data.osm_keys_tags() {
+    build_msg DATA "update searx/data/osm_keys_tags.json"
+    pyenv.cmd python searx_extra/update/update_osm_keys_tags.py
+    dump_return $?
+}
+
 docs.prebuild() {
     build_msg DOCS "build ${DOCS_BUILD}/includes"
     (

searx/data/__init__.py

@@ -1,25 +1,47 @@
+# SPDX-License-Identifier: AGPL-3.0-or-later
+# lint: pylint
+"""This module holds the *data* created by::
+
+  make data.all
+
+"""
+
+__all__ = [
+    'ENGINES_LANGUAGES',
+    'CURRENCIES',
+    'USER_AGENTS',
+    'EXTERNAL_URLS',
+    'WIKIDATA_UNITS',
+    'EXTERNAL_BANGS',
+    'OSM_KEYS_TAGS',
+    'ahmia_blacklist_loader',
+]
+
 import json
 from pathlib import Path
 
-
-__init__ = ['ENGINES_LANGUGAGES', 'CURRENCIES', 'USER_AGENTS', 'EXTERNAL_URLS', 'WIKIDATA_UNITS', 'EXTERNAL_BANGS',
-            'bangs_loader', 'ahmia_blacklist_loader']
 data_dir = Path(__file__).parent
 
-
+def _load(filename):
-def load(filename):
+    with open(data_dir / filename, encoding='utf-8') as f:
-    with open(data_dir / filename, encoding='utf-8') as fd:
+        return json.load(f)
-        return json.load(fd)
-
 
 def ahmia_blacklist_loader():
-    with open(str(data_dir / 'ahmia_blacklist.txt'), encoding='utf-8') as fd:
+    """Load data from `ahmia_blacklist.txt` and return a list of MD5 values of onion
-        return fd.read().split()
+    names.  The MD5 values are fetched by::
 
+      searx_extra/update/update_ahmia_blacklist.py
 
-ENGINES_LANGUAGES = load('engines_languages.json')
+    This function is used by :py:mod:`searx.plugins.ahmia_filter`.
-CURRENCIES = load('currencies.json')
+
-USER_AGENTS = load('useragents.json')
+    """
-EXTERNAL_URLS = load('external_urls.json')
+    with open(str(data_dir / 'ahmia_blacklist.txt'), encoding='utf-8') as f:
-WIKIDATA_UNITS = load('wikidata_units.json')
+        return f.read().split()
-EXTERNAL_BANGS = load('external_bangs.json')
+
+ENGINES_LANGUAGES = _load('engines_languages.json')
+CURRENCIES = _load('currencies.json')
+USER_AGENTS = _load('useragents.json')
+EXTERNAL_URLS = _load('external_urls.json')
+WIKIDATA_UNITS = _load('wikidata_units.json')
+EXTERNAL_BANGS = _load('external_bangs.json')
+OSM_KEYS_TAGS = _load('osm_keys_tags.json')

searx/engines/openstreetmap.py

@@ -1,12 +1,22 @@
 # SPDX-License-Identifier: AGPL-3.0-or-later
+# lint: pylint
+"""OpenStreetMap (Map)
+
 """
- OpenStreetMap (Map)
+# pylint: disable=missing-function-docstring
-"""
 
 import re
 from json import loads
+from urllib.parse import urlencode
+from functools import partial
+
 from flask_babel import gettext
 
+from searx.data import OSM_KEYS_TAGS, CURRENCIES
+from searx.utils import searx_useragent
+from searx.external_urls import get_external_url
+from searx.engines.wikidata import send_wikidata_query, sparql_string_escape
+
 # about
 about = {
     "website": 'https://www.openstreetmap.org/',
@@ -23,89 +33,409 @@ paging = False
 
 # search-url
 base_url = 'https://nominatim.openstreetmap.org/'
-search_string = 'search/{query}?format=json&polygon_geojson=1&addressdetails=1'
+search_string = 'search?{query}&polygon_geojson=1&format=jsonv2&addressdetails=1&extratags=1&dedupe=1'
-result_base_url = 'https://openstreetmap.org/{osm_type}/{osm_id}'
+result_id_url = 'https://openstreetmap.org/{osm_type}/{osm_id}'
+result_lat_lon_url = 'https://www.openstreetmap.org/?mlat={lat}&mlon={lon}&zoom={zoom}&layers=M'
 
-route_url = 'https://graphhopper.com/maps/?point={}&point={}&locale=en-US&vehicle=car&weighting=fastest&turn_costs=true&use_miles=false&layer=Omniscale'  # noqa
+route_url = 'https://graphhopper.com/maps/?point={}&point={}&locale=en-US&vehicle=car&weighting=fastest&turn_costs=true&use_miles=false&layer=Omniscale'  # pylint: disable=line-too-long
 route_re = re.compile('(?:from )?(.+) to (.+)')
 
+wikidata_image_sparql = """
+select ?item ?itemLabel ?image ?sign ?symbol ?website ?wikipediaName
+where {
+  values ?item { %WIKIDATA_IDS% }
+  OPTIONAL { ?item wdt:P18|wdt:P8517|wdt:P4291|wdt:P5252|wdt:P3451|wdt:P4640|wdt:P5775|wdt:P2716|wdt:P1801|wdt:P4896 ?image }
+  OPTIONAL { ?item wdt:P1766|wdt:P8505|wdt:P8667 ?sign }
+  OPTIONAL { ?item wdt:P41|wdt:P94|wdt:P154|wdt:P158|wdt:P2910|wdt:P4004|wdt:P5962|wdt:P8972 ?symbol }
+  OPTIONAL { ?item wdt:P856 ?website }
+  SERVICE wikibase:label {
+    bd:serviceParam wikibase:language "%LANGUAGE%,en".
+    ?item rdfs:label ?itemLabel .
+  }
+  OPTIONAL {
+    ?wikipediaUrl schema:about ?item;
+                  schema:isPartOf/wikibase:wikiGroup "wikipedia";
+                  schema:name ?wikipediaName;
+                  schema:inLanguage "%LANGUAGE%" .
+  }
+}
+ORDER by ?item
+"""
+
+
+# key value that are link: mapping functions
+# 'mapillary': P1947
+# but https://github.com/kartaview/openstreetcam.org/issues/60
+# but https://taginfo.openstreetmap.org/keys/kartaview ...
+def value_to_https_link(value):
+    http = 'http://'
+    if value.startswith(http):
+        value = 'https://' + value[len(http) :]
+    return (value, value)
+
+
+def value_to_website_link(value):
+    value = value.split(';')[0]
+    return (value, value)
+
+
+def value_wikipedia_link(value):
+    value = value.split(':', 1)
+    return ('https://{0}.wikipedia.org/wiki/{1}'.format(*value), '{1} ({0})'.format(*value))
+
+
+def value_with_prefix(prefix, value):
+    return (prefix + value, value)
+
+
+VALUE_TO_LINK = {
+    'website': value_to_website_link,
+    'contact:website': value_to_website_link,
+    'email': partial(value_with_prefix, 'mailto:'),
+    'contact:email': partial(value_with_prefix, 'mailto:'),
+    'contact:phone': partial(value_with_prefix, 'tel:'),
+    'phone': partial(value_with_prefix, 'tel:'),
+    'fax': partial(value_with_prefix, 'fax:'),
+    'contact:fax': partial(value_with_prefix, 'fax:'),
+    'contact:mastodon': value_to_https_link,
+    'facebook': value_to_https_link,
+    'contact:facebook': value_to_https_link,
+    'contact:foursquare': value_to_https_link,
+    'contact:instagram': value_to_https_link,
+    'contact:linkedin': value_to_https_link,
+    'contact:pinterest': value_to_https_link,
+    'contact:telegram': value_to_https_link,
+    'contact:tripadvisor': value_to_https_link,
+    'contact:twitter': value_to_https_link,
+    'contact:yelp': value_to_https_link,
+    'contact:youtube': value_to_https_link,
+    'contact:webcam': value_to_website_link,
+    'wikipedia': value_wikipedia_link,
+    'wikidata': partial(value_with_prefix, 'https://wikidata.org/wiki/'),
+    'brand:wikidata': partial(value_with_prefix, 'https://wikidata.org/wiki/'),
+}
+KEY_ORDER = [
+    'cuisine',
+    'organic',
+    'delivery',
+    'delivery:covid19',
+    'opening_hours',
+    'opening_hours:covid19',
+    'fee',
+    'payment:*',
+    'currency:*',
+    'outdoor_seating',
+    'bench',
+    'wheelchair',
+    'level',
+    'building:levels',
+    'bin',
+    'public_transport',
+    'internet_access:ssid',
+]
+KEY_RANKS = {k: i for i, k in enumerate(KEY_ORDER)}
+
 
-# do search-request
 def request(query, params):
-
+    """do search-request"""
-    params['url'] = base_url + search_string.format(query=query)
+    params['url'] = base_url + search_string.format(query=urlencode({'q': query}))
     params['route'] = route_re.match(query)
-
+    params['headers']['User-Agent'] = searx_useragent()
     return params
 
 
-# get response from search-request
 def response(resp):
+    """get response from search-request"""
     results = []
-    json = loads(resp.text)
+    nominatim_json = loads(resp.text)
+    user_language = resp.search_params['language']
 
     if resp.search_params['route']:
         results.append({
             'answer': gettext('Get directions'),
             'url': route_url.format(*resp.search_params['route'].groups()),
-        })
+            })
 
-    # parse results
+    fetch_wikidata(nominatim_json, user_language)
-    for r in json:
+
-        if 'display_name' not in r:
+    for result in nominatim_json:
+        title, address = get_title_address(result)
+
+        # ignore result without title
+        if not title:
             continue
 
-        title = r['display_name'] or ''
+        url, osm, geojson = get_url_osm_geojson(result)
-        osm_type = r.get('osm_type', r.get('type'))
+        img_src = get_img_src(result)
-        url = result_base_url.format(osm_type=osm_type,
+        links, link_keys = get_links(result, user_language)
-                                     osm_id=r['osm_id'])
+        data = get_data(result, user_language, link_keys)
 
-        osm = {'type': osm_type,
+        results.append({
-               'id': r['osm_id']}
+            'template': 'map.html',
+            'title': title,
+            'address': address,
+            'address_label': get_key_label('addr', user_language),
+            'url': url,
+            'osm': osm,
+            'geojson': geojson,
+            'img_src': img_src,
+            'links': links,
+            'data': data,
+            'type': get_tag_label(
+                result.get('category'), result.get('type', ''), user_language
+            ),
+            'type_icon': result.get('icon'),
+            'content': '',
+            'longitude': result['lon'],
+            'latitude': result['lat'],
+            'boundingbox': result['boundingbox'],
+        })
 
-        geojson = r.get('geojson')
-
-        # if no geojson is found and osm_type is a node, add geojson Point
-        if not geojson and osm_type == 'node':
-            geojson = {'type': 'Point', 'coordinates': [r['lon'], r['lat']]}
-
-        address_raw = r.get('address')
-        address = {}
-
-        # get name
-        if r['class'] == 'amenity' or\
-           r['class'] == 'shop' or\
-           r['class'] == 'tourism' or\
-           r['class'] == 'leisure':
-            if address_raw.get('address29'):
-                address = {'name': address_raw.get('address29')}
-            else:
-                address = {'name': address_raw.get(r['type'])}
-
-        # add rest of adressdata, if something is already found
-        if address.get('name'):
-            address.update({'house_number': address_raw.get('house_number'),
-                           'road': address_raw.get('road'),
-                           'locality': address_raw.get('city',
-                                       address_raw.get('town',          # noqa
-                                       address_raw.get('village'))),    # noqa
-                           'postcode': address_raw.get('postcode'),
-                           'country': address_raw.get('country'),
-                           'country_code': address_raw.get('country_code')})
-        else:
-            address = None
-
-        # append result
-        results.append({'template': 'map.html',
-                        'title': title,
-                        'content': '',
-                        'longitude': r['lon'],
-                        'latitude': r['lat'],
-                        'boundingbox': r['boundingbox'],
-                        'geojson': geojson,
-                        'address': address,
-                        'osm': osm,
-                        'url': url})
-
-    # return results
     return results
+
+
+def get_wikipedia_image(raw_value):
+    if not raw_value:
+        return None
+    return get_external_url('wikimedia_image', raw_value)
+
+
+def fetch_wikidata(nominatim_json, user_langage):
+    """Update nominatim_json using the result of an unique to wikidata
+
+    For result in nominatim_json:
+        If result['extratags']['wikidata'] or r['extratags']['wikidata link']:
+            Set result['wikidata'] to { 'image': ..., 'image_sign':..., 'image_symbal':... }
+            Set result['extratags']['wikipedia'] if not defined
+            Set result['extratags']['contact:website'] if not defined
+    """
+    wikidata_ids = []
+    wd_to_results = {}
+    for result in nominatim_json:
+        e = result.get("extratags")
+        if e:
+            # ignore brand:wikidata
+            wd_id = e.get("wikidata", e.get("wikidata link"))
+            if wd_id and wd_id not in wikidata_ids:
+                wikidata_ids.append("wd:" + wd_id)
+                wd_to_results.setdefault(wd_id, []).append(result)
+
+    if wikidata_ids:
+        wikidata_ids_str = " ".join(wikidata_ids)
+        query = wikidata_image_sparql.replace('%WIKIDATA_IDS%', sparql_string_escape(wikidata_ids_str)).replace(
+            '%LANGUAGE%', sparql_string_escape(user_langage)
+        )
+        wikidata_json = send_wikidata_query(query)
+        for wd_result in wikidata_json.get('results', {}).get('bindings', {}):
+            wd_id = wd_result['item']['value'].replace('http://www.wikidata.org/entity/', '')
+            for result in wd_to_results.get(wd_id, []):
+                result['wikidata'] = {
+                    'itemLabel': wd_result['itemLabel']['value'],
+                    'image': get_wikipedia_image(wd_result.get('image', {}).get('value')),
+                    'image_sign': get_wikipedia_image(wd_result.get('sign', {}).get('value')),
+                    'image_symbol': get_wikipedia_image(wd_result.get('symbol', {}).get('value')),
+                }
+                # overwrite wikipedia link
+                wikipedia_name = wd_result.get('wikipediaName', {}).get('value')
+                if wikipedia_name:
+                    result['extratags']['wikipedia'] = user_langage + ':' + wikipedia_name
+                # get website if not already defined
+                website = wd_result.get('website', {}).get('value')
+                if (
+                    website
+                    and not result['extratags'].get('contact:website')
+                    and not result['extratags'].get('website')
+                ):
+                    result['extratags']['contact:website'] = website
+
+
+def get_title_address(result):
+    """Return title and address
+
+    title may be None
+    """
+    address_raw = result.get('address')
+    address_name = None
+    address = {}
+
+    # get name
+    if (
+        result['category'] == 'amenity'
+        or result['category'] == 'shop'
+        or result['category'] == 'tourism'
+        or result['category'] == 'leisure'
+    ):
+        if address_raw.get('address29'):
+            # https://github.com/osm-search/Nominatim/issues/1662
+            address_name = address_raw.get('address29')
+        else:
+            address_name =  address_raw.get(result['category'])
+    elif result['type'] in address_raw:
+        address_name =  address_raw.get(result['type'])
+
+    # add rest of adressdata, if something is already found
+    if address_name:
+        title = address_name
+        address.update(
+            {
+                'name': address_name,
+                'house_number': address_raw.get('house_number'),
+                'road': address_raw.get('road'),
+                'locality': address_raw.get(
+                    'city', address_raw.get('town', address_raw.get('village'))  # noqa
+                ),  # noqa
+                'postcode': address_raw.get('postcode'),
+                'country': address_raw.get('country'),
+                'country_code': address_raw.get('country_code'),
+            }
+        )
+    else:
+        title = result.get('display_name')
+
+    return title, address
+
+
+def get_url_osm_geojson(result):
+    """Get url, osm and geojson
+    """
+    osm_type = result.get('osm_type', result.get('type'))
+    if 'osm_id' not in result:
+        # see https://github.com/osm-search/Nominatim/issues/1521
+        # query example: "EC1M 5RF London"
+        url = result_lat_lon_url.format(lat=result['lat'], lon=result['lon'], zoom=12)
+        osm = {}
+    else:
+        url = result_id_url.format(osm_type=osm_type, osm_id=result['osm_id'])
+        osm = {'type': osm_type, 'id': result['osm_id']}
+
+    geojson = result.get('geojson')
+    # if no geojson is found and osm_type is a node, add geojson Point
+    if not geojson and osm_type == 'node':
+        geojson = {'type': 'Point', 'coordinates': [result['lon'], result['lat']]}
+
+    return url, osm, geojson
+
+
+def get_img_src(result):
+    """Get image URL from either wikidata or r['extratags']"""
+    # wikidata
+    img_src = None
+    if 'wikidata' in result:
+        img_src = result['wikidata']['image']
+        if not img_src:
+            img_src = result['wikidata']['image_symbol']
+        if not img_src:
+            img_src = result['wikidata']['image_sign']
+
+    # img_src
+    if not img_src and result.get('extratags', {}).get('image'):
+        img_src = result['extratags']['image']
+        del result['extratags']['image']
+    if not img_src and result.get('extratags', {}).get('wikimedia_commons'):
+        img_src = get_external_url('wikimedia_image', result['extratags']['wikimedia_commons'])
+        del result['extratags']['wikimedia_commons']
+
+    return img_src
+
+
+def get_links(result, user_language):
+    """Return links from result['extratags']"""
+    links = []
+    link_keys = set()
+    for k, mapping_function in VALUE_TO_LINK.items():
+        raw_value = result['extratags'].get(k)
+        if raw_value:
+            url, url_label = mapping_function(raw_value)
+            if url.startswith('https://wikidata.org'):
+                url_label = result.get('wikidata', {}).get('itemLabel') or url_label
+            links.append({
+                'label': get_key_label(k, user_language),
+                'url': url,
+                'url_label': url_label,
+            })
+            link_keys.add(k)
+    return links, link_keys
+
+
+def get_data(result, user_language, ignore_keys):
+    """Return key, value of result['extratags']
+
+    Must be call after get_links
+
+    Note: the values are not translated
+    """
+    data = []
+    for k, v in result['extratags'].items():
+        if k in ignore_keys:
+            continue
+        if get_key_rank(k) is None:
+            continue
+        k_label = get_key_label(k, user_language)
+        if k_label:
+            data.append({
+                'label': k_label,
+                'key': k,
+                'value': v,
+            })
+    data.sort(key=lambda entry: (get_key_rank(entry['key']), entry['label']))
+    return data
+
+
+def get_key_rank(k):
+    """Get OSM key rank
+
+    The rank defines in which order the key are displayed in the HTML result
+    """
+    key_rank = KEY_RANKS.get(k)
+    if key_rank is None:
+        # "payment:*" in KEY_ORDER matches "payment:cash", "payment:debit card", etc...
+        key_rank = KEY_RANKS.get(k.split(':')[0] + ':*')
+    return key_rank
+
+
+def get_label(labels, lang):
+    """Get label from labels in OSM_KEYS_TAGS
+
+    in OSM_KEYS_TAGS, labels have key == '*'
+    """
+    tag_label = labels.get(lang.lower())
+    if tag_label is None:
+        # example: if 'zh-hk' is not found, check 'zh'
+        tag_label = labels.get(lang.split('-')[0])
+    if tag_label is None and lang != 'en':
+        # example: if 'zh' is not found, check 'en'
+        tag_label = labels.get('en')
+    if tag_label is None and len(labels.values()) > 0:
+        # example: if still not found, use the first entry
+        tag_label = labels.values()[0]
+    return tag_label
+
+
+def get_tag_label(tag_category, tag_name, lang):
+    """Get tag label from OSM_KEYS_TAGS"""
+    tag_name = '' if tag_name is None else tag_name
+    tag_labels = OSM_KEYS_TAGS['tags'].get(tag_category, {}).get(tag_name, {})
+    return get_label(tag_labels, lang)
+
+
+def get_key_label(key_name, lang):
+    """Get key label from OSM_KEYS_TAGS"""
+    if key_name.startswith('currency:'):
+        # currency:EUR --> get the name from the CURRENCIES variable
+        # see https://wiki.openstreetmap.org/wiki/Key%3Acurrency
+        # and for exampe https://taginfo.openstreetmap.org/keys/currency:EUR#values
+        # but there is also currency=EUR (currently not handled)
+        # https://taginfo.openstreetmap.org/keys/currency#values
+        currency = key_name.split(':')
+        if len(currency) > 1:
+            o = CURRENCIES['iso4217'].get(currency)
+            if o:
+                return get_label(o, lang).lower()
+            return currency
+
+    labels = OSM_KEYS_TAGS['keys']
+    for k in key_name.split(':') + ['*']:
+        labels = labels.get(k)
+        if labels is None:
+            return None
+    return get_label(labels, lang)

searx/external_urls.py

@@ -10,6 +10,7 @@ IMDB_PREFIX_TO_URL_ID = {
     'co': 'imdb_company',
     'ev': 'imdb_event'
 }
+HTTP_WIKIMEDIA_IMAGE = 'http://commons.wikimedia.org/wiki/Special:FilePath/'
 
 
 def get_imdb_url_id(imdb_item_id):
@@ -17,6 +18,14 @@ def get_imdb_url_id(imdb_item_id):
     return IMDB_PREFIX_TO_URL_ID.get(id_prefix)
 
 
+def get_wikimedia_image_id(url):
+    if url.startswith(HTTP_WIKIMEDIA_IMAGE):
+        return url[len(HTTP_WIKIMEDIA_IMAGE):]
+    if url.startswith('File:'):
+        return url[len('File:'):]
+    return url
+
+
 def get_external_url(url_id, item_id, alternative="default"):
     """Return an external URL or None if url_id is not found.
 
@@ -25,8 +34,11 @@ def get_external_url(url_id, item_id, alternative="default"):
 
     If item_id is None, the raw URL with the $1 is returned.
     """
-    if url_id == 'imdb_id' and item_id is not None:
+    if item_id is not None:
-        url_id = get_imdb_url_id(item_id)
+        if url_id == 'imdb_id':
+            url_id = get_imdb_url_id(item_id)
+        elif url_id == 'wikimedia_image':
+            item_id = get_wikimedia_image_id(item_id)
 
     url_description = EXTERNAL_URLS.get(url_id)
     if url_description:

searx/static/themes/oscar/css/logicodev-dark.css

@@ -623,6 +623,30 @@ input[type=checkbox]:not(:checked) + .label_hide_if_checked + .label_hide_if_not
 .result-map {
   clear: both;
 }
+.result-map .img-thumbnail {
+  float: right;
+  width: auto;
+  height: 120px;
+  border: 0;
+  background: inherit;
+}
+.result-map .img-type {
+  width: 20px;
+  max-height: 20px;
+}
+.result-map .result-map-details {
+  font-size: 13px;
+  border-collapse: separate;
+  border-spacing: 0 0.35rem;
+}
+.result-map .result-map-details th {
+  font-weight: inherit;
+  width: 20rem;
+  vertical-align: top;
+}
+.result-map .result-map-details td {
+  vertical-align: top;
+}
 .result-code {
   clear: both;
 }

searx/static/themes/oscar/css/logicodev.css

@@ -650,6 +650,30 @@ input[type=checkbox]:not(:checked) + .label_hide_if_checked + .label_hide_if_not
 .result-map {
   clear: both;
 }
+.result-map .img-thumbnail {
+  float: right;
+  width: auto;
+  height: 120px;
+  border: 0;
+  background: inherit;
+}
+.result-map .img-type {
+  width: 20px;
+  max-height: 20px;
+}
+.result-map .result-map-details {
+  font-size: 13px;
+  border-collapse: separate;
+  border-spacing: 0 0.35rem;
+}
+.result-map .result-map-details th {
+  font-weight: inherit;
+  width: 20rem;
+  vertical-align: top;
+}
+.result-map .result-map-details td {
+  vertical-align: top;
+}
 .result-code {
   clear: both;
 }

searx/static/themes/oscar/src/less/logicodev/results.less

@@ -171,6 +171,36 @@
 // map formating of results
 .result-map {
     clear: both;
+
+    .img-thumbnail {
+        float: right;
+        width: auto;
+        height: 120px;
+        border: 0;
+        background: inherit;
+    }
+
+    .img-type {
+        width: 20px;
+        max-height: 20px;
+    }
+
+    .result-map-details {
+        font-size: 13px;
+        border-collapse: separate;
+        border-spacing: 0 0.35rem;
+
+        th {
+            font-weight: inherit;
+            width: 20rem;
+            vertical-align: top;
+        }
+
+        td {
+            vertical-align: top;
+        }
+    }
+
 }
 
 // code formating of results

searx/static/themes/simple/css/searx-rtl.css

@@ -1,4 +1,4 @@
-/*! searx | 01-06-2021 | https://github.com/searxng/searxng */
+/*! searx | 09-06-2021 | https://github.com/searxng/searxng */
 /*
 * searx, A privacy-respecting, hackable metasearch engine
 *
@@ -2346,6 +2346,30 @@ article.result-images[data-vim-selected]::before {
   background-color: rgba(0, 0, 0, 0.6);
   font-size: 0.7em;
 }
+.result-map img.image {
+  float: right !important;
+  height: 100px !important;
+  width: auto !important;
+}
+.result-map table {
+  font-size: 0.9em;
+  width: auto;
+  border-collapse: separate;
+  border-spacing: 0 0.35rem;
+}
+.result-map table th {
+  font-weight: inherit;
+  width: 17rem;
+  vertical-align: top;
+  text-align: left;
+}
+.result-map table td {
+  vertical-align: top;
+  text-align: left;
+}
+.hidden {
+  display: none !important;
+}
 .torrent_result {
   border-left: 10px solid lightgray;
   padding-left: 3px;

searx/static/themes/simple/css/searx.css

@@ -1,4 +1,4 @@
-/*! searx | 01-06-2021 | https://github.com/searxng/searxng */
+/*! searx | 09-06-2021 | https://github.com/searxng/searxng */
 /*
 * searx, A privacy-respecting, hackable metasearch engine
 *
@@ -2346,6 +2346,30 @@ article.result-images[data-vim-selected]::before {
   background-color: rgba(0, 0, 0, 0.6);
   font-size: 0.7em;
 }
+.result-map img.image {
+  float: right !important;
+  height: 100px !important;
+  width: auto !important;
+}
+.result-map table {
+  font-size: 0.9em;
+  width: auto;
+  border-collapse: separate;
+  border-spacing: 0 0.35rem;
+}
+.result-map table th {
+  font-weight: inherit;
+  width: 17rem;
+  vertical-align: top;
+  text-align: left;
+}
+.result-map table td {
+  vertical-align: top;
+  text-align: left;
+}
+.hidden {
+  display: none !important;
+}
 .torrent_result {
   border-left: 10px solid lightgray;
   padding-left: 3px;

searx/static/themes/simple/js/searx.head.min.js

@@ -1,4 +1,4 @@
-/*! simple/searx.min.js | 01-06-2021 | https://github.com/searxng/searxng */
+/*! simple/searx.min.js | 09-06-2021 | https://github.com/searxng/searxng */
 
 (function(t,e){"use strict";var a=e.currentScript||function(){var t=e.getElementsByTagName("script");return t[t.length-1]}();t.searx={touch:"ontouchstart"in t||t.DocumentTouch&&document instanceof DocumentTouch||false,method:a.getAttribute("data-method"),autocompleter:a.getAttribute("data-autocompleter")==="true",search_on_category_select:a.getAttribute("data-search-on-category-select")==="true",infinite_scroll:a.getAttribute("data-infinite-scroll")==="true",static_path:a.getAttribute("data-static-path"),translations:JSON.parse(a.getAttribute("data-translations"))};e.getElementsByTagName("html")[0].className=t.searx.touch?"js touch":"js"})(window,document);
 //# sourceMappingURL=searx.head.min.js.map

searx/static/themes/simple/js/searx.min.js

@@ -1,4 +1,4 @@
-/*! simple/searx.min.js | 01-06-2021 | https://github.com/searxng/searxng */
+/*! simple/searx.min.js | 09-06-2021 | https://github.com/searxng/searxng */
 
 window.searx=function(t,o){"use strict";if(t.Element){(function(e){e.matches=e.matches||e.matchesSelector||e.webkitMatchesSelector||e.msMatchesSelector||function(e){var t=this,n=(t.parentNode||t.document).querySelectorAll(e),i=-1;while(n[++i]&&n[i]!=t);return!!n[i]}})(Element.prototype)}function a(e,t,n){try{e.call(t,n)}catch(e){console.log(e)}}var s=window.searx||{};s.on=function(i,e,r,t){t=t||false;if(typeof i!=="string"){i.addEventListener(e,r,t)}else{o.addEventListener(e,function(e){var t=e.target||e.srcElement,n=false;while(t&&t.matches&&t!==o&&!(n=t.matches(i)))t=t.parentElement;if(n)a(r,t,e)},t)}};s.ready=function(e){if(document.readyState!="loading"){e.call(t)}else{t.addEventListener("DOMContentLoaded",e.bind(t))}};s.http=function(e,t,n){var i=new XMLHttpRequest,r=function(){},o=function(){},a={then:function(e){r=e;return a},catch:function(e){o=e;return a}};try{i.open(e,t,true);i.onload=function(){if(i.status==200){r(i.response,i.responseType)}else{o(Error(i.statusText))}};i.onerror=function(){o(Error("Network Error"))};i.onabort=function(){o(Error("Transaction is aborted"))};i.send()}catch(e){o(e)}return a};s.loadStyle=function(e){var t=s.static_path+e,n="style_"+e.replace(".","_"),i=o.getElementById(n);if(i===null){i=o.createElement("link");i.setAttribute("id",n);i.setAttribute("rel","stylesheet");i.setAttribute("type","text/css");i.setAttribute("href",t);o.body.appendChild(i)}};s.loadScript=function(e,t){var n=s.static_path+e,i="script_"+e.replace(".","_"),r=o.getElementById(i);if(r===null){r=o.createElement("script");r.setAttribute("id",i);r.setAttribute("src",n);r.onload=t;r.onerror=function(){r.setAttribute("error","1")};o.body.appendChild(r)}else if(!r.hasAttribute("error")){try{t.apply(r,[])}catch(e){console.log(e)}}else{console.log("callback not executed : script '"+n+"' not loaded.")}};s.insertBefore=function(e,t){element.parentNode.insertBefore(e,t)};s.insertAfter=function(e,t){t.parentNode.insertBefore(e,t.nextSibling)};s.on(".close","click",function(e){var t=e.target||e.srcElement;this.parentNode.classList.add("invisible")});return s}(window,document);(function(e){if(typeof exports==="object"&&typeof module!=="undefined"){module.exports=e()}else if(typeof define==="function"&&define.amd){define([],e)}else{var t;if(typeof window!=="undefined"){t=window}else if(typeof global!=="undefined"){t=global}else if(typeof self!=="undefined"){t=self}else{t=this}t.AutoComplete=e()}})(function(){var e,t,n;return function o(a,s,l){function u(n,e){if(!s[n]){if(!a[n]){var t=typeof require=="function"&&require;if(!e&&t)return t(n,!0);if(c)return c(n,!0);var i=new Error("Cannot find module '"+n+"'");throw i.code="MODULE_NOT_FOUND",i}var r=s[n]={exports:{}};a[n][0].call(r.exports,function(e){var t=a[n][1][e];return u(t?t:e)},r,r.exports,o,a,s,l)}return s[n].exports}var c=typeof require=="function"&&require;for(var e=0;e<l.length;e++)u(l[e]);return u}({1:[function(e,t,n){
 /*

searx/static/themes/simple/less/style.less

@@ -304,6 +304,38 @@ article.result-images[data-vim-selected]::before {
   }
 }
 
+.result-map {
+
+  img.image {
+    float: right !important;
+    height: 100px !important;
+    width: auto !important;
+  }
+
+  table {
+    font-size: .9em;
+    width: auto;
+    border-collapse: separate;
+    border-spacing: 0 0.35rem;
+
+    th {
+      font-weight: inherit;
+      width: 17rem;
+      vertical-align: top;
+      text-align: left;
+    }
+
+    td {
+      vertical-align: top;
+      text-align: left;
+    }
+  }
+}
+
+.hidden {
+  display: none !important;
+}
+
 .torrent_result {
   border-left: 10px solid @color-result-torrent-border;
   padding-left: 3px;

searx/templates/oscar/result_templates/map.html

@@ -3,48 +3,64 @@
 {{- result_header(result, favicons, loop.index) -}}
 {{- result_sub_header(result, loop.index) -}}
 
-{%- if (result.latitude and result.longitude) or result.boundingbox -%}
-    <small> &bull; <a class="text-info btn-collapse collapsed searx_init_map cursor-pointer disabled_if_nojs" data-toggle="collapse" data-target="#result-map-{{ index }}" data-leaflet-target="osm-map-{{ index }}" data-map-lon="{{ result.longitude }}" data-map-lat="{{ result.latitude }}" {% if result.boundingbox %}data-map-boundingbox='{{ result.boundingbox|tojson|safe }}'{% endif %} {% if result.geojson %}data-map-geojson='{{ result.geojson|tojson|safe }}'{% endif %} data-btn-text-collapsed="{{ _('show map') }}" data-btn-text-not-collapsed="{{ _('hide map') }}">{{ icon('globe') }} {{ _('show map') }}</a></small>
-{%- endif -%}
-
-{# {% if (result.latitude and result.longitude) %}
-    <small> &bull; <a class="text-info btn-collapse collapsed cursor-pointer disabled_if_nojs" data-toggle="collapse" data-target="#result-geodata-{{ index }}" data-btn-text-collapsed="{{ _('show geodata') }}" data-btn-text-not-collapsed="{{ _('hide geodata') }}">{{ icon('map-marker') }} {{ _('show geodata') }}</a></small>
-{% endif %} #}
-
 <div class="container-fluid">
 
-{%- if result.address -%}
+    <div class="row">
-<p class="row result-content result-adress col-xs-12 col-sm-5 col-md-4" itemscope itemtype="http://schema.org/PostalAddress">
+        {%- if result.img_src -%}<img src="{{ image_proxify(result.img_src) }}" title="{{ result.title|striptags }}" class="img-thumbnail">{%- endif -%}
-    {%- if result.address.name -%}
+
-        <strong itemprop="name">{{ result.address.name }}</strong><br/>
+        {%- if result.type_icon or result.type -%}
-    {%- endif -%}
+        <p class="result-content">
-    {%- if result.address.road -%}
+            {%- if result.type_icon -%}
-        <span itemprop="streetAddress">
+                <img src="{{ image_proxify(result.type_icon) }}" title="{{ result.type|striptags }}" class="img-type">
-            {%- if result.address.house_number %}{{ result.address.house_number }}, {% endif -%}
+            {%- endif -%}
-            {{- result.address.road -}}
+            {{ result.type if result.type is not none else '' }}
-        </span><br/>
+        </p>
-    {%- endif -%}
+        {%- endif -%}
-    {%- if result.address.locality -%}
+
-        <span itemprop="addressLocality">{{ result.address.locality }}</span>
+        {%- if result.content %}<p class="result-content">{{ result.content|safe }}</p>{% endif -%}
-        {%- if result.address.postcode %}, <span itemprop="postalCode">{{ result.address.postcode }}</span>{% endif -%}
+
-        <br/>
+        <table class="result-content result-map-details">
-    {%- endif -%}
+            {%- if result.address -%}
-    {%- if result.address.country -%}
+            <tr>
-        <span itemprop="addressCountry">{{ result.address.country }}</span>
+                <th scope="row">
-    {%- endif -%}
+                    {{ result.address_label or _('address') }}
-</p>
+                </th>
-{%- endif %}
+                <td class="result-content result-adress" itemscope itemtype="http://schema.org/PostalAddress">
+                {%- if result.address.name -%}
+                    <strong itemprop="name" class="hidden">{{ result.address.name }}</strong>
+                {%- endif -%}
+                {%- if result.address.road -%}
+                    <span itemprop="streetAddress">
+                        {%- if result.address.house_number %}{{ result.address.house_number }}, {% endif -%}
+                        {{- result.address.road -}}
+                    </span><br/>
+                {%- endif -%}
+                {%- if result.address.locality -%}
+                    <span itemprop="addressLocality">{{ result.address.locality }}</span>
+                    {%- if result.address.postcode %}, <span itemprop="postalCode">{{ result.address.postcode }}</span>{% endif -%}
+                    <br/>
+                {%- endif -%}
+                {%- if result.address.country -%}
+                    <span itemprop="addressCountry">{{ result.address.country }}</span>
+                {%- endif -%}
+                </td>
+            </tr>
+            {%- endif %}
+            {%- for info in result.data -%}
+            <tr><th scope="row">{{ info.label }}</th><td>{{ info.value|safe }}</td></tr>
+            {%- endfor -%}
+            {%- for link in result.links -%}
+            <tr><th scope="row">{{ link.label }}</th><td><a class="text-info cursor-pointer" href="{{ link.url }}">{{ link.url_label|safe }}</a></td></tr>
+            {%- endfor -%}
+        </table>
+
+        <p class="result-content">
+            {%- if (result.latitude and result.longitude) or result.boundingbox -%}
+                <span><a class="text-info btn-collapse collapsed searx_init_map cursor-pointer disabled_if_nojs" data-toggle="collapse" data-target="#result-map-{{ index }}" data-leaflet-target="osm-map-{{ index }}" data-map-lon="{{ result.longitude }}" data-map-lat="{{ result.latitude }}" {% if result.boundingbox %}data-map-boundingbox='{{ result.boundingbox|tojson|safe }}'{% endif %} {% if result.geojson %}data-map-geojson='{{ result.geojson|tojson|safe }}'{% endif %} data-btn-text-collapsed="{{ _('show map') }}" data-btn-text-not-collapsed="{{ _('hide map') }}">{{ icon('globe') }} {{ _('show map') }}</a></span>
+            {%- endif -%}
+        </p>
 
-{# {% if (result.latitude and result.longitude) %}
-    <div class="row collapse col-xs-12 col-sm-5 col-md-4" id="result-geodata-{{ index }}">
-        <strong>Longitude:</strong> {{ result.longitude }} <br/>
-        <strong>Latitude:</strong> {{ result.latitude }}
     </div>
-{% endif %} #}
-
-{%- if result.content %}<p class="row result-content col-xs-12 col-sm-12 col-md-12">{{ result.content|safe }}</p>{% endif -%}
-
-</div>
 
 {%- if (result.latitude and result.longitude) or result.boundingbox -%}
     <div class="collapse" id="result-map-{{ index }}">
@@ -52,6 +68,7 @@
     </div>
 {%- endif -%}
 
+</div>
 {%- if rtl -%}
 {{- result_footer_rtl(result, loop.index) -}}
 {% else %}

searx/templates/simple/result_templates/map.html

@@ -3,50 +3,46 @@
 {{ result_header(result, favicons, image_proxify) -}}
 {{- result_sub_header(result) -}}
 
-{%- if (result.latitude and result.longitude) or result.boundingbox -%}
-    <small> &bull; <a class="btn-collapse collapsed searx_init_map hide_if_nojs" data-target="#result-map-{{ index }}" data-btn-text-collapsed="{{ _('show map') }}" data-btn-text-not-collapsed="{{ _('hide map') }}" data-leaflet-target="osm-map-{{ index }}" data-map-lon="{{ result.longitude }}" data-map-lat="{{ result.latitude }}" {% if result.boundingbox %}data-map-boundingbox='{{ result.boundingbox|tojson|safe }}'{% endif %} {% if result.geojson %}data-map-geojson='{{ result.geojson|tojson|safe }}'{% endif %}>{{ icon( 'globe') }} {{ _('show map') }}</a></small>
-{%- endif -%}
-
-{#-
-{% if (result.latitude and result.longitude) %}
-    <small> &bull; <a class="btn-collapse disabled_if_nojs" data-target="#result-geodata-{{ index }}" data-btn-text-collapsed="{{ _('show geodata') }}" data-btn-text-not-collapsed="{{ _('hide geodata') }}">{{ icon('location') }} {{ _('show geodata') }}</a></small>
-{% endif %}
--#}
-
-{%- if result.address -%}
-<p itemscope itemtype="http://schema.org/PostalAddress" class="content">
-    {%- if result.address.name -%}
-        <strong itemprop="name">{{ result.address.name }}</strong><br/>
-    {%- endif -%}
-    {% if result.address.road -%}
-        <span itemprop="streetAddress">
-            {%- if result.address.house_number -%}{{- result.address.house_number -}}, {% endif %}
-            {{- result.address.road -}}
-        </span><br/>
-    {%- endif %}
-    {%- if result.address.locality -%}
-        <span itemprop="addressLocality">{{- result.address.locality -}}</span>
-        {%- if result.address.postcode -%}, <span itemprop="postalCode">{{- result.address.postcode -}}</span>{% endif %}
-        <br/>
-    {%- endif -%}
-    {%- if result.address.country -%}
-        <span itemprop="addressCountry">{{- result.address.country -}}</span>
-    {%- endif -%}
-</p>
-{%- endif -%}
-
 {%- if result.content %}<p class="content">{{ result.content|safe }}</p>{% endif -%}
 
-{{- result_sub_footer(result, proxify) -}}
+<table>
+    {%- if result.address -%}
+    <tr>
+        <th scope="row">{{ result.address_label or _('address') }}</th>
+        <td itemscope itemtype="http://schema.org/PostalAddress">
+            {%- if result.address.name -%}
+                <strong itemprop="name" class="hidden">{{ result.address.name }}</strong>
+            {%- endif -%}
+            {% if result.address.road -%}
+                <span itemprop="streetAddress">
+                    {%- if result.address.house_number -%}{{- result.address.house_number -}}, {% endif %}
+                    {{- result.address.road -}}
+                </span><br/>
+            {%- endif %}
+            {%- if result.address.locality -%}
+                <span itemprop="addressLocality">{{- result.address.locality -}}</span>
+                {%- if result.address.postcode -%}, <span itemprop="postalCode">{{- result.address.postcode -}}</span>{% endif %}
+                <br/>
+            {%- endif -%}
+            {%- if result.address.country -%}
+                <span itemprop="addressCountry">{{- result.address.country -}}</span>
+            {%- endif -%}
+        </td>
+    </tr>
+    {%- endif %}
+    {%- for info in result.data -%}
+    <tr><th scope="row">{{ info.label }}</th><td>{{ info.value|safe }}</td></tr>
+    {%- endfor -%}
+    {%- for link in result.links -%}
+    <tr><th scope="row">{{ link.label }}</th><td><a class="text-info cursor-pointer" href="{{ link.url }}">{{ link.url|safe }}</a></td></tr>
+    {%- endfor -%}
+</table>
 
-{#-
+{%- if (result.latitude and result.longitude) or result.boundingbox -%}
-{% if (result.latitude and result.longitude) -%}
+    <small> <a class="btn-collapse collapsed searx_init_map hide_if_nojs" data-target="#result-map-{{ index }}" data-btn-text-collapsed="{{ _('show map') }}" data-btn-text-not-collapsed="{{ _('hide map') }}" data-leaflet-target="osm-map-{{ index }}" data-map-lon="{{ result.longitude }}" data-map-lat="{{ result.latitude }}" {% if result.boundingbox %}data-map-boundingbox='{{ result.boundingbox|tojson|safe }}'{% endif %} {% if result.geojson %}data-map-geojson='{{ result.geojson|tojson|safe }}'{% endif %}>{{ icon( 'globe') }} {{ _('show map') }}</a></small>
-    <div id="result-geodata-{{ index }}" class="invisible">
+{%- endif -%}
-        <strong>Longitude:</strong> {{ result.longitude }} <br/>
+
-        <strong>Latitude:</strong> {{ result.latitude }}
+{{- result_sub_footer(result, proxify) -}}
-    </div>
-{%- endif %}
--#}
 
 {% if (result.latitude and result.longitude) or result.boundingbox -%}
     <div id="result-map-{{ index }}" class="invisible"><div id="osm-map-{{ index }}" style="height:300px; width:100%; margin: 10px 0;" ></div></div>

searx/utils.py

@@ -45,7 +45,7 @@ def searx_useragent():
     """Return the searx User Agent"""
     return 'searx/{searx_version} {suffix}'.format(
            searx_version=VERSION_STRING,
-           suffix=settings['outgoing']['useragent_suffix'].strip())
+           suffix=settings['outgoing']['useragent_suffix']).strip()
 
 
 def gen_useragent(os=None):

searx_extra/update/update_osm_keys_tags.py

@@ -0,0 +1,209 @@
+#!/usr/bin/env python
+# lint: pylint
+# pylint: disable=missing-function-docstring
+"""Fetch OSM keys and tags.
+
+To get the i18n names, the scripts uses `Wikidata Query Service`_ instead of for
+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``).
+
+"""
+
+import json
+import collections
+from pathlib import Path
+
+from searx import searx_dir
+from searx.network import set_timeout_for_thread
+from searx.engines.wikidata import send_wikidata_query
+from searx.languages import language_codes
+from searx.engines.openstreetmap import get_key_rank, VALUE_TO_LINK
+
+SPARQL_TAGS_REQUEST = """
+SELECT ?tag ?item ?itemLabel WHERE {
+  ?item wdt:P1282 ?tag .
+  ?item rdfs:label ?itemLabel .
+  FILTER(STRSTARTS(?tag, 'Tag'))
+}
+GROUP BY ?tag ?item ?itemLabel
+ORDER BY ?tag ?item ?itemLabel
+"""
+
+SPARQL_KEYS_REQUEST = """
+SELECT ?key ?item ?itemLabel WHERE {
+  ?item wdt:P1282 ?key .
+  ?item rdfs:label ?itemLabel .
+  FILTER(STRSTARTS(?key, 'Key'))
+}
+GROUP BY ?key ?item ?itemLabel
+ORDER BY ?key ?item ?itemLabel
+"""
+
+LANGUAGES = [l[0].lower() for l in language_codes]
+
+PRESET_KEYS = {
+    ('wikidata',): {'en': 'Wikidata'},
+    ('wikipedia',): {'en': 'Wikipedia'},
+    ('email',): {'en': 'Email'},
+    ('facebook',): {'en': 'Facebook'},
+    ('fax',): {'en': 'Fax'},
+    ('internet_access', 'ssid'): {'en': 'Wi-Fi'},
+}
+
+INCLUDED_KEYS = {
+    ('addr', )
+}
+
+def get_preset_keys():
+    results = collections.OrderedDict()
+    for keys, value in PRESET_KEYS.items():
+        r = results
+        for k in keys:
+            r = r.setdefault(k, {})
+        r.setdefault('*', value)
+    return results
+
+def get_keys():
+    results = get_preset_keys()
+    response = send_wikidata_query(SPARQL_KEYS_REQUEST)
+
+    for key in response['results']['bindings']:
+        keys = key['key']['value'].split(':')[1:]
+        if keys[0] == 'currency' and len(keys) > 1:
+            # special case in openstreetmap.py
+            continue
+        if keys[0] == 'contact' and len(keys) > 1:
+            # label for the key "contact.email" is "Email"
+            # whatever the language
+            r = results.setdefault('contact', {})
+            r[keys[1]] = {
+                '*': {
+                    'en': keys[1]
+                }
+            }
+            continue
+        if tuple(keys) in PRESET_KEYS:
+            # skip presets (already set above)
+            continue
+        if get_key_rank(':'.join(keys)) is None\
+            and ':'.join(keys) not in VALUE_TO_LINK\
+            and tuple(keys) not in INCLUDED_KEYS:
+            # keep only keys that will be displayed by openstreetmap.py
+            continue
+        label = key['itemLabel']['value'].lower()
+        lang = key['itemLabel']['xml:lang']
+        r = results
+        for k in keys:
+            r = r.setdefault(k, {})
+        r = r.setdefault('*', {})
+        if lang in LANGUAGES:
+            r.setdefault(lang, label)
+
+    # special cases
+    results['delivery']['covid19']['*'].clear()
+    for k, v in results['delivery']['*'].items():
+        results['delivery']['covid19']['*'][k] = v + ' (COVID19)'
+
+    results['opening_hours']['covid19']['*'].clear()
+    for k, v in results['opening_hours']['*'].items():
+        results['opening_hours']['covid19']['*'][k] = v + ' (COVID19)'
+
+    return results
+
+
+def get_tags():
+    results = collections.OrderedDict()
+    response = send_wikidata_query(SPARQL_TAGS_REQUEST)
+    for tag in response['results']['bindings']:
+        tag_names = tag['tag']['value'].split(':')[1].split('=')
+        if len(tag_names) == 2:
+            tag_category, tag_type = tag_names
+        else:
+            tag_category, tag_type = tag_names[0], ''
+        label = tag['itemLabel']['value'].lower()
+        lang = tag['itemLabel']['xml:lang']
+        if lang in LANGUAGES:
+            results.setdefault(tag_category, {}).setdefault(tag_type, {}).setdefault(lang, label)
+    return results
+
+def optimize_data_lang(translations):
+    language_to_delete = []
+    # remove "zh-hk" entry if the value is the same as "zh"
+    # same for "en-ca" / "en" etc...
+    for language in translations:
+        if '-' in language:
+            base_language = language.split('-')[0]
+            if translations.get(base_language) == translations.get(language):
+                language_to_delete.append(language)
+
+    for language in language_to_delete:
+        del translations[language]
+    language_to_delete = []
+
+    # remove entries that have the same value than the "en" entry
+    value_en = translations.get('en')
+    if value_en:
+        for language, value in translations.items():
+            if language != 'en' and value == value_en:
+                language_to_delete.append(language)
+
+    for language in language_to_delete:
+        del translations[language]
+
+def optimize_tags(data):
+    for v in data.values():
+        for translations in v.values():
+            optimize_data_lang(translations)
+    return data
+
+def optimize_keys(data):
+    for k, v in data.items():
+        if k == '*':
+            optimize_data_lang(v)
+        elif isinstance(v, dict):
+            optimize_keys(v)
+    return data
+
+def get_osm_tags_filename():
+    return Path(searx_dir) / "data" / "osm_keys_tags.json"
+
+if __name__ == '__main__':
+
+    set_timeout_for_thread(60)
+    result = {
+        'keys': optimize_keys(get_keys()),
+        'tags': optimize_tags(get_tags()),
+    }
+    with open(get_osm_tags_filename(), 'w') as f:
+        json.dump(result, f, indent=4, ensure_ascii=False)