Merge pull request #1866 from return42/fix-news

bugfix: google-news and bing-news has changed the language parameter

Makefile

@@ -27,6 +27,7 @@ help:
 	@echo  '  uninstall - uninstall (./local)'
 	@echo  '  gh-pages  - build docs & deploy on gh-pages branch'
 	@echo  '  clean     - drop builds and environments'
+	@echo  '  project   - re-build generic files of the searx project'
 	@echo  ''
 	@$(MAKE) -s -f utils/makefile.include make-help
 	@echo  ''
@@ -67,6 +68,18 @@ docs-live:  pyenvinstall sphinx-live
 $(GH_PAGES)::
 	@echo "doc available at --> $(DOCS_URL)"
 
+# update project files
+# --------------------
+
+PHONY += project engines-languages
+
+project: searx/data/engines_languages.json
+
+searx/data/engines_languages.json:  pyenvinstall
+	$(PY_ENV_ACT); python utils/fetch_languages.py
+	mv engines_languages.json searx/data/engines_languages.json
+	mv languages.py searx/languages.py
+
 # test
 # ----
 

requirements-dev.txt

@@ -5,6 +5,7 @@ mock==2.0.0
 nose2[coverage_plugin]
 cov-core==1.15.0
 pep8==1.7.0
+pylint
 plone.testing==5.0.0
 splinter==0.11.0
 transifex-client==0.12.2

searx/engines/bing.py

@@ -110,13 +110,18 @@ def response(resp):
 
 # get supported languages from their site
 def _fetch_supported_languages(resp):
-    supported_languages = []
-    dom = html.fromstring(resp.text)
-    options = eval_xpath(dom, '//div[@id="limit-languages"]//input')
-    for option in options:
-        code = eval_xpath(option, './@id')[0].replace('_', '-')
-        if code == 'nb':
-            code = 'no'
-        supported_languages.append(code)
+    lang_tags = set()
 
-    return supported_languages
+    setmkt = re.compile('setmkt=([^&]*)')
+    dom = html.fromstring(resp.text)
+    lang_links = eval_xpath(dom, "//li/a[contains(@href, 'setmkt')]")
+
+    for a in lang_links:
+        href = eval_xpath(a, './@href')[0]
+        match = setmkt.search(href)
+        l_tag = match.groups()[0]
+        _lang, _nation = l_tag.split('-', 1)
+        l_tag = _lang.lower() + '-' + _nation.upper()
+        lang_tags.add(l_tag)
+
+    return list(lang_tags)

searx/engines/bing_images.py

@@ -18,6 +18,8 @@ import re
 from searx.url_utils import urlencode
 from searx.utils import match_language
 
+from searx.engines.bing import _fetch_supported_languages, supported_languages_url, language_aliases
+
 # engine dependent config
 categories = ['images']
 paging = True
@@ -103,22 +105,3 @@ def response(resp):
             continue
 
     return results
-
-
-# get supported languages from their site
-def _fetch_supported_languages(resp):
-    supported_languages = []
-    dom = html.fromstring(resp.text)
-
-    regions_xpath = '//div[@id="region-section-content"]' \
-                    + '//ul[@class="b_vList"]/li/a/@href'
-
-    regions = dom.xpath(regions_xpath)
-    for region in regions:
-        code = re.search('setmkt=[^\&]+', region).group()[7:]
-        if code == 'nb-NO':
-            code = 'no-NO'
-
-        supported_languages.append(code)
-
-    return supported_languages

searx/engines/bing_news.py

@@ -15,9 +15,10 @@ from datetime import datetime
 from dateutil import parser
 from lxml import etree
 from searx.utils import list_get, match_language
-from searx.engines.bing import _fetch_supported_languages, supported_languages_url, language_aliases
 from searx.url_utils import urlencode, urlparse, parse_qsl
 
+from searx.engines.bing import _fetch_supported_languages, supported_languages_url, language_aliases
+
 # engine dependent config
 categories = ['news']
 paging = True
@@ -58,6 +59,7 @@ def _get_url(query, language, offset, time_range):
             offset=offset,
             interval=time_range_dict[time_range])
     else:
+        # e.g. setmkt=de-de&setlang=de
         search_path = search_string.format(
             query=urlencode({'q': query, 'setmkt': language}),
             offset=offset)

searx/engines/bing_videos.py

@@ -12,10 +12,10 @@
 
 from json import loads
 from lxml import html
-from searx.engines.bing_images import _fetch_supported_languages, supported_languages_url
 from searx.url_utils import urlencode
 from searx.utils import match_language
 
+from searx.engines.bing import _fetch_supported_languages, supported_languages_url, language_aliases
 
 categories = ['videos']
 paging = True
@@ -67,6 +67,10 @@ def request(query, params):
     if params['time_range'] in time_range_dict:
         params['url'] += time_range_string.format(interval=time_range_dict[params['time_range']])
 
+    # bing videos did not like "older" versions < 70.0.1 when selectin other
+    # languages then 'en' .. very strange ?!?!
+    params['headers']['User-Agent'] = 'Mozilla/5.0 (X11; Linux x86_64; rv:73.0.1) Gecko/20100101 Firefox/73.0.1'
+
     return params
 
 

searx/engines/google_news.py

@@ -54,7 +54,7 @@ def request(query, params):
     if params['language'] != 'all':
         language = match_language(params['language'], supported_languages, language_aliases).split('-')[0]
         if language:
-            params['url'] += '&lr=lang_' + language
+            params['url'] += '&hl=' + language
 
     return params
 

searx/languages.py

@@ -3,9 +3,11 @@
 # this file is generated automatically by utils/update_search_languages.py
 
 language_codes = (
+    (u"af-NA", u"Afrikaans", u"", u"Afrikaans"),
     (u"ar-SA", u"العربية", u"", u"Arabic"),
+    (u"be-BY", u"Беларуская", u"", u"Belarusian"),
     (u"bg-BG", u"Български", u"", u"Bulgarian"),
-    (u"ca-ES", u"Català", u"", u"Catalan"),
+    (u"ca-AD", u"Català", u"", u"Catalan"),
     (u"cs-CZ", u"Čeština", u"", u"Czech"),
     (u"da-DK", u"Dansk", u"", u"Danish"),
     (u"de", u"Deutsch", u"", u"German"),
@@ -17,11 +19,15 @@ language_codes = (
     (u"en-AU", u"English", u"Australia", u"English"),
     (u"en-CA", u"English", u"Canada", u"English"),
     (u"en-GB", u"English", u"United Kingdom", u"English"),
+    (u"en-IE", u"English", u"Ireland", u"English"),
     (u"en-IN", u"English", u"India", u"English"),
-    (u"en-MY", u"English", u"Malaysia", u"English"),
+    (u"en-NZ", u"English", u"New Zealand", u"English"),
+    (u"en-PH", u"English", u"Philippines", u"English"),
+    (u"en-SG", u"English", u"Singapore", u"English"),
     (u"en-US", u"English", u"United States", u"English"),
     (u"es", u"Español", u"", u"Spanish"),
     (u"es-AR", u"Español", u"Argentina", u"Spanish"),
+    (u"es-CL", u"Español", u"Chile", u"Spanish"),
     (u"es-ES", u"Español", u"España", u"Spanish"),
     (u"es-MX", u"Español", u"México", u"Spanish"),
     (u"et-EE", u"Eesti", u"", u"Estonian"),
@@ -35,6 +41,7 @@ language_codes = (
     (u"he-IL", u"עברית", u"", u"Hebrew"),
     (u"hr-HR", u"Hrvatski", u"", u"Croatian"),
     (u"hu-HU", u"Magyar", u"", u"Hungarian"),
+    (u"hy-AM", u"Հայերեն", u"", u"Armenian"),
     (u"id-ID", u"Indonesia", u"", u"Indonesian"),
     (u"is-IS", u"Íslenska", u"", u"Icelandic"),
     (u"it-IT", u"Italiano", u"", u"Italian"),
@@ -42,7 +49,7 @@ language_codes = (
     (u"ko-KR", u"한국어", u"", u"Korean"),
     (u"lt-LT", u"Lietuvių", u"", u"Lithuanian"),
     (u"lv-LV", u"Latviešu", u"", u"Latvian"),
-    (u"ms-MY", u"Bahasa Melayu", u"", u"Malay"),
+    (u"ms-MY", u"Melayu", u"", u"Malay"),
     (u"nb-NO", u"Norsk Bokmål", u"", u"Norwegian Bokmål"),
     (u"nl", u"Nederlands", u"", u"Dutch"),
     (u"nl-BE", u"Nederlands", u"België", u"Dutch"),
@@ -55,8 +62,9 @@ language_codes = (
     (u"ru-RU", u"Русский", u"", u"Russian"),
     (u"sk-SK", u"Slovenčina", u"", u"Slovak"),
     (u"sl-SI", u"Slovenščina", u"", u"Slovenian"),
-    (u"sr-RS", u"Српски", u"", u"Serbian"),
+    (u"sr-RS", u"Srpski", u"", u"Serbian"),
     (u"sv-SE", u"Svenska", u"", u"Swedish"),
+    (u"sw-KE", u"Kiswahili", u"", u"Swahili"),
     (u"th-TH", u"ไทย", u"", u"Thai"),
     (u"tr-TR", u"Türkçe", u"", u"Turkish"),
     (u"uk-UA", u"Українська", u"", u"Ukrainian"),

utils/fetch_languages.py

@@ -5,7 +5,7 @@
 # Output files (engines_languages.json and languages.py)
 # are written in current directory to avoid overwriting in case something goes wrong.
 
-from json import dump
+import json
 import io
 from sys import path
 from babel import Locale, UnknownLocaleError
@@ -22,19 +22,22 @@ languages_file = 'languages.py'
 
 # Fetchs supported languages for each engine and writes json file with those.
 def fetch_supported_languages():
+
     engines_languages = {}
-    for engine_name in engines:
+    names = list(engines)
+    names.sort()
+
+    for engine_name in names:
+        print("fetching languages of engine %s" % engine_name)
+
         if hasattr(engines[engine_name], 'fetch_supported_languages'):
-            try:
-                engines_languages[engine_name] = engines[engine_name].fetch_supported_languages()
-                if type(engines_languages[engine_name]) == list:
-                    engines_languages[engine_name] = sorted(engines_languages[engine_name])
-            except Exception as e:
-                print(e)
+            engines_languages[engine_name] = engines[engine_name].fetch_supported_languages()
+            if type(engines_languages[engine_name]) == list:
+                engines_languages[engine_name] = sorted(engines_languages[engine_name])
 
     # write json file
-    with io.open(engines_languages_file, "w", encoding="utf-8") as f:
-        dump(engines_languages, f, ensure_ascii=False, indent=4, separators=(',', ': '))
+    with open(engines_languages_file, 'w', encoding='utf-8') as f:
+        json.dump(engines_languages, f, indent=2, sort_keys=True)
 
     return engines_languages