Cutlery/searxng
1
0
Fork 0
mirror of https://github.com/searxng/searxng.git synced 2025-11-26 00:05:08 -05:00
Code Issues Packages Projects Releases Wiki Activity

[fix] brave engines - web, images & videos (#5478)

Browse Source 
brave web:
  xpath selectors needed to be justified

brave images & videos:
  The JS code with the JS object was read incorrectly; not always, but quite
  often, it led to exceptions when the Python data structure was created from it.

BTW: A complete review was conducted and corrections or additions were made to
the type definitions.

To test all brave engines in once::

    !br !brimg !brvid !brnews weather

Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
This commit is contained in:
Markus Heiser 2025-11-25 13:28:47 +01:00 committed by GitHub
parent 54a97e1043
commit ebb9ea4571
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 153 additions and 163 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

307
searx/engines/brave.py
View File

@ -124,17 +124,17 @@ from urllib.parse import (
urlparse, urlparse,
) )
import json
from dateutil import parser from dateutil import parser
from lxml import html from lxml import html
from searx import locales from searx import locales
from searx.utils import ( from searx.utils import (
extr,
extract_text, extract_text,
eval_xpath,
eval_xpath_list, eval_xpath_list,
eval_xpath_getindex, eval_xpath_getindex,
js_obj_str_to_python, js_obj_str_to_python,
js_obj_str_to_json_str,
get_embeded_stream_url, get_embeded_stream_url,
) )
from searx.enginelib.traits import EngineTraits from searx.enginelib.traits import EngineTraits
@ -142,17 +142,17 @@ from searx.result_types import EngineResults
from searx.extended_types import SXNG_Response from searx.extended_types import SXNG_Response
about = { about = {
"website": 'https://search.brave.com/', "website": "https://search.brave.com/",
"wikidata_id": 'Q22906900', "wikidata_id": "Q22906900",
"official_api_documentation": None, "official_api_documentation": None,
"use_official_api": False, "use_official_api": False,
"require_api_key": False, "require_api_key": False,
"results": 'HTML', "results": "HTML",
} }
base_url = "https://search.brave.com/" base_url = "https://search.brave.com/"
categories = [] categories = []
brave_category: t.Literal["search", "videos", "images", "news", "goggles"] = 'search' brave_category: t.Literal["search", "videos", "images", "news", "goggles"] = "search"
"""Brave supports common web-search, videos, images, news, and goggles search. """Brave supports common web-search, videos, images, news, and goggles search.
- ``search``: Common WEB search - ``search``: Common WEB search
@ -182,71 +182,86 @@ to do more won't return any result and you will most likely be flagged as a bot.
""" """
safesearch = True safesearch = True
safesearch_map = {2: 'strict', 1: 'moderate', 0: 'off'} # cookie: safesearch=off safesearch_map = {2: "strict", 1: "moderate", 0: "off"} # cookie: safesearch=off
time_range_support = False time_range_support = False
"""Brave only supports time-range in :py:obj:`brave_category` ``search`` (UI """Brave only supports time-range in :py:obj:`brave_category` ``search`` (UI
category All) and in the goggles category.""" category All) and in the goggles category."""
time_range_map: dict[str, str] = { time_range_map: dict[str, str] = {
'day': 'pd', "day": "pd",
'week': 'pw', "week": "pw",
'month': 'pm', "month": "pm",
'year': 'py', "year": "py",
} }
def request(query: str, params: dict[str, t.Any]) -> None: def request(query: str, params: dict[str, t.Any]) -> None:
args: dict[str, t.Any] = { args: dict[str, t.Any] = {
'q': query, "q": query,
'source': 'web', "source": "web",
} }
if brave_spellcheck: if brave_spellcheck:
args['spellcheck'] = '1' args["spellcheck"] = "1"
if brave_category in ('search', 'goggles'): if brave_category in ("search", "goggles"):
if params.get('pageno', 1) - 1: if params.get("pageno", 1) - 1:
args['offset'] = params.get('pageno', 1) - 1 args["offset"] = params.get("pageno", 1) - 1
if time_range_map.get(params['time_range']): if time_range_map.get(params["time_range"]):
args['tf'] = time_range_map.get(params['time_range']) args["tf"] = time_range_map.get(params["time_range"])
if brave_category == 'goggles': if brave_category == "goggles":
args['goggles_id'] = Goggles args["goggles_id"] = Goggles
params["url"] = f"{base_url}{brave_category}?{urlencode(args)}" params["url"] = f"{base_url}{brave_category}?{urlencode(args)}"
logger.debug("url %s", params["url"])
# set properties in the cookies # set properties in the cookies
params['cookies']['safesearch'] = safesearch_map.get(params['safesearch'], 'off') params["cookies"]["safesearch"] = safesearch_map.get(params["safesearch"], "off")
# the useLocation is IP based, we use cookie 'country' for the region # the useLocation is IP based, we use cookie "country" for the region
params['cookies']['useLocation'] = '0' params["cookies"]["useLocation"] = "0"
params['cookies']['summarizer'] = '0' params["cookies"]["summarizer"] = "0"
engine_region = traits.get_region(params['searxng_locale'], 'all') engine_region = traits.get_region(params["searxng_locale"], "all")
params['cookies']['country'] = engine_region.split('-')[-1].lower() # type: ignore params["cookies"]["country"] = engine_region.split("-")[-1].lower() # type: ignore
ui_lang = locales.get_engine_locale(params['searxng_locale'], traits.custom["ui_lang"], 'en-us') ui_lang = locales.get_engine_locale(params["searxng_locale"], traits.custom["ui_lang"], "en-us")
params['cookies']['ui_lang'] = ui_lang params["cookies"]["ui_lang"] = ui_lang
logger.debug("cookies %s", params["cookies"])
logger.debug("cookies %s", params['cookies'])
params['headers']['Sec-Fetch-Dest'] = "document"
params['headers']['Sec-Fetch-Mode'] = "navigate"
params['headers']['Sec-Fetch-Site'] = "same-origin"
params['headers']['Sec-Fetch-User'] = "?1"
def _extract_published_date(published_date_raw): def _extract_published_date(published_date_raw: str | None):
if published_date_raw is None: if published_date_raw is None:
return None return None
try: try:
return parser.parse(published_date_raw) return parser.parse(published_date_raw)
except parser.ParserError: except parser.ParserError:
return None return None
def extract_json_data(text: str) -> dict[str, t.Any]:
# Example script source containing the data:
#
# kit.start(app, element, {
# node_ids: [0, 19],
# data: [{type:"data",data: .... ["q","goggles_id"],route:1,url:1}}]
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
text = text[text.index("<script") : text.index("</script")]
if not text:
raise ValueError("can't find JS/JSON data in the given text")
start = text.index("data: [{")
end = text.rindex("}}]")
js_obj_str = text[start:end]
js_obj_str = "{" + js_obj_str + "}}]}"
# js_obj_str = js_obj_str.replace("\xa0", "") # remove ASCII for &nbsp;
# js_obj_str = js_obj_str.replace(r"\u003C", "<").replace(r"\u003c", "<") # fix broken HTML tags in strings
json_str = js_obj_str_to_json_str(js_obj_str)
data: dict[str, t.Any] = json.loads(json_str)
return data
def response(resp: SXNG_Response) -> EngineResults: def response(resp: SXNG_Response) -> EngineResults:
if brave_category in ('search', 'goggles'): if brave_category in ('search', 'goggles'):
@ -261,11 +276,8 @@ def response(resp: SXNG_Response) -> EngineResults:
# node_ids: [0, 19], # node_ids: [0, 19],
# data: [{type:"data",data: .... ["q","goggles_id"],route:1,url:1}}] # data: [{type:"data",data: .... ["q","goggles_id"],route:1,url:1}}]
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ # ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
js_object = "[{" + extr(resp.text, "data: [{", "}}],") + "}}]" json_data: dict[str, t.Any] = extract_json_data(resp.text)
json_data = js_obj_str_to_python(js_object) json_resp: dict[str, t.Any] = json_data['data'][1]["data"]['body']['response']
# json_data is a list and at the second position (0,1) in this list we find the "response" data we need ..
json_resp = json_data[1]['data']['body']['response']
if brave_category == 'images': if brave_category == 'images':
return _parse_images(json_resp) return _parse_images(json_resp)
@ -275,150 +287,121 @@ def response(resp: SXNG_Response) -> EngineResults:
raise ValueError(f"Unsupported brave category: {brave_category}") raise ValueError(f"Unsupported brave category: {brave_category}")
def _parse_search(resp) -> EngineResults: def _parse_search(resp: SXNG_Response) -> EngineResults:
result_list = EngineResults() res = EngineResults()
dom = html.fromstring(resp.text) dom = html.fromstring(resp.text)
# I doubt that Brave is still providing the "answer" class / I haven't seen for result in eval_xpath_list(dom, "//div[contains(@class, 'snippet ')]"):
# answers in brave for a long time.
answer_tag = eval_xpath_getindex(dom, '//div[@class="answer"]', 0, default=None)
if answer_tag:
url = eval_xpath_getindex(dom, '//div[@id="featured_snippet"]/a[@class="result-header"]/@href', 0, default=None)
answer = extract_text(answer_tag)
if answer is not None:
result_list.add(result_list.types.Answer(answer=answer, url=url))
# xpath_results = '//div[contains(@class, "snippet fdb") and @data-type="web"]' url: str | None = eval_xpath_getindex(result, ".//a/@href", 0, default=None)
xpath_results = '//div[contains(@class, "snippet ")]' title_tag = eval_xpath_getindex(result, ".//div[contains(@class, 'title')]", 0, default=None)
for result in eval_xpath_list(dom, xpath_results):
url = eval_xpath_getindex(result, './/a[contains(@class, "h")]/@href', 0, default=None)
title_tag = eval_xpath_getindex(
result, './/a[contains(@class, "h")]//div[contains(@class, "title")]', 0, default=None
)
if url is None or title_tag is None or not urlparse(url).netloc: # partial url likely means it's an ad if url is None or title_tag is None or not urlparse(url).netloc: # partial url likely means it's an ad
continue continue
content: str = extract_text( content: str = ""
eval_xpath_getindex(result, './/div[contains(@class, "snippet-description")]', 0, default='') pub_date = None
) # type: ignore
pub_date_raw = eval_xpath(result, 'substring-before(.//div[contains(@class, "snippet-description")], "-")')
pub_date = _extract_published_date(pub_date_raw)
if pub_date and content.startswith(pub_date_raw):
content = content.lstrip(pub_date_raw).strip("- \n\t")
thumbnail = eval_xpath_getindex(result, './/img[contains(@class, "thumb")]/@src', 0, default='') _content = eval_xpath_getindex(result, ".//div[contains(@class, 'content')]", 0, default="")
if len(_content):
content = extract_text(_content) # type: ignore
_pub_date = extract_text(
eval_xpath_getindex(_content, ".//span[contains(@class, 't-secondary')]", 0, default="")
)
if _pub_date:
pub_date = _extract_published_date(_pub_date)
content = content.lstrip(_pub_date).strip("- \n\t")
item = { thumbnail: str = eval_xpath_getindex(result, ".//a[contains(@class, 'thumbnail')]//img/@src", 0, default="")
'url': url,
'title': extract_text(title_tag), item = res.types.LegacyResult(
'content': content, template="default.html",
'publishedDate': pub_date, url=url,
'thumbnail': thumbnail, title=extract_text(title_tag),
} content=content,
publishedDate=pub_date,
thumbnail=thumbnail,
)
res.add(item)
video_tag = eval_xpath_getindex( video_tag = eval_xpath_getindex(
result, './/div[contains(@class, "video-snippet") and @data-macro="video"]', 0, default=None result, ".//div[contains(@class, 'video-snippet') and @data-macro='video']", 0, default=[]
) )
if video_tag is not None: if len(video_tag):
# In my tests a video tag in the WEB search was most often not a # In my tests a video tag in the WEB search was most often not a
# video, except the ones from youtube .. # video, except the ones from youtube ..
iframe_src = get_embeded_stream_url(url) iframe_src = get_embeded_stream_url(url)
if iframe_src: if iframe_src:
item['iframe_src'] = iframe_src item["iframe_src"] = iframe_src
item['template'] = 'videos.html' item["template"] = "videos.html"
item['thumbnail'] = eval_xpath_getindex(video_tag, './/img/@src', 0, default='')
pub_date_raw = extract_text(
eval_xpath(video_tag, './/div[contains(@class, "snippet-attributes")]/div/text()')
)
item['publishedDate'] = _extract_published_date(pub_date_raw)
else:
item['thumbnail'] = eval_xpath_getindex(video_tag, './/img/@src', 0, default='')
result_list.append(item) return res
return result_list
def _parse_news(resp) -> EngineResults: def _parse_news(resp: SXNG_Response) -> EngineResults:
res = EngineResults()
result_list = EngineResults()
dom = html.fromstring(resp.text) dom = html.fromstring(resp.text)
for result in eval_xpath_list(dom, '//div[contains(@class, "results")]//div[@data-type="news"]'): for result in eval_xpath_list(dom, "//div[contains(@class, 'results')]//div[@data-type='news']"):
# import pdb url = eval_xpath_getindex(result, ".//a[contains(@class, 'result-header')]/@href", 0, default=None)
# pdb.set_trace()
url = eval_xpath_getindex(result, './/a[contains(@class, "result-header")]/@href', 0, default=None)
if url is None: if url is None:
continue continue
title = extract_text(eval_xpath_list(result, './/span[contains(@class, "snippet-title")]')) title = eval_xpath_list(result, ".//span[contains(@class, 'snippet-title')]")
content = extract_text(eval_xpath_list(result, './/p[contains(@class, "desc")]')) content = eval_xpath_list(result, ".//p[contains(@class, 'desc')]")
thumbnail = eval_xpath_getindex(result, './/div[contains(@class, "image-wrapper")]//img/@src', 0, default='') thumbnail = eval_xpath_getindex(result, ".//div[contains(@class, 'image-wrapper')]//img/@src", 0, default="")
item = { item = res.types.LegacyResult(
"url": url, template="default.html",
"title": title, url=url,
"content": content, title=extract_text(title),
"thumbnail": thumbnail, thumbnail=thumbnail,
} content=extract_text(content),
)
res.add(item)
result_list.append(item) return res
return result_list
def _parse_images(json_resp) -> EngineResults: def _parse_images(json_resp: dict[str, t.Any]) -> EngineResults:
result_list = EngineResults() res = EngineResults()
for result in json_resp["results"]: for result in json_resp["results"]:
item = { item = res.types.LegacyResult(
'url': result['url'], template="images.html",
'title': result['title'], url=result["url"],
'content': result['description'], title=result["title"],
'template': 'images.html', source=result["source"],
'resolution': result['properties']['format'], img_src=result["properties"]["url"],
'source': result['source'], thumbnail_src=result["thumbnail"]["src"],
'img_src': result['properties']['url'], )
'thumbnail_src': result['thumbnail']['src'], res.add(item)
}
result_list.append(item)
return result_list return res
def _parse_videos(json_resp) -> EngineResults: def _parse_videos(json_resp: dict[str, t.Any]) -> EngineResults:
result_list = EngineResults() res = EngineResults()
for result in json_resp["results"]: for result in json_resp["results"]:
item = res.types.LegacyResult(
url = result['url'] template="videos.html",
item = { url=result["url"],
'url': url, title=result["title"],
'title': result['title'], content=result["description"],
'content': result['description'], length=result["video"]["duration"],
'template': 'videos.html', duration=result["video"]["duration"],
'length': result['video']['duration'], publishedDate=_extract_published_date(result["age"]),
'duration': result['video']['duration'], )
'publishedDate': _extract_published_date(result['age']), if result["thumbnail"] is not None:
} item["thumbnail"] = result["thumbnail"]["src"]
iframe_src = get_embeded_stream_url(result["url"])
if result['thumbnail'] is not None:
item['thumbnail'] = result['thumbnail']['src']
iframe_src = get_embeded_stream_url(url)
if iframe_src: if iframe_src:
item['iframe_src'] = iframe_src item["iframe_src"] = iframe_src
result_list.append(item) res.add(item)
return result_list return res
def fetch_traits(engine_traits: EngineTraits): def fetch_traits(engine_traits: EngineTraits):
@ -443,21 +426,21 @@ def fetch_traits(engine_traits: EngineTraits):
print("ERROR: response from Brave is not OK.") print("ERROR: response from Brave is not OK.")
dom = html.fromstring(resp.text) dom = html.fromstring(resp.text)
for option in dom.xpath('//section//option[@value="en-us"]/../option'): for option in dom.xpath("//section//option[@value='en-us']/../option"):
ui_lang = option.get('value') ui_lang = option.get("value")
try: try:
l = babel.Locale.parse(ui_lang, sep='-') l = babel.Locale.parse(ui_lang, sep="-")
if l.territory: if l.territory:
sxng_tag = region_tag(babel.Locale.parse(ui_lang, sep='-')) sxng_tag = region_tag(babel.Locale.parse(ui_lang, sep="-"))
else: else:
sxng_tag = language_tag(babel.Locale.parse(ui_lang, sep='-')) sxng_tag = language_tag(babel.Locale.parse(ui_lang, sep="-"))
except babel.UnknownLocaleError: except babel.UnknownLocaleError:
print("ERROR: can't determine babel locale of Brave's (UI) language %s" % ui_lang) print("ERROR: can't determine babel locale of Brave's (UI) language %s" % ui_lang)
continue continue
conflict = engine_traits.custom["ui_lang"].get(sxng_tag) conflict = engine_traits.custom["ui_lang"].get(sxng_tag) # type: ignore
if conflict: if conflict:
if conflict != ui_lang: if conflict != ui_lang:
print("CONFLICT: babel %s --> %s, %s" % (sxng_tag, conflict, ui_lang)) print("CONFLICT: babel %s --> %s, %s" % (sxng_tag, conflict, ui_lang))
@ -466,26 +449,26 @@ def fetch_traits(engine_traits: EngineTraits):
# search regions of brave # search regions of brave
resp = get('https://cdn.search.brave.com/serp/v2/_app/immutable/chunks/parameters.734c106a.js') resp = get("https://cdn.search.brave.com/serp/v2/_app/immutable/chunks/parameters.734c106a.js")
if not resp.ok: if not resp.ok:
print("ERROR: response from Brave is not OK.") print("ERROR: response from Brave is not OK.")
country_js = resp.text[resp.text.index("options:{all") + len('options:') :] country_js = resp.text[resp.text.index("options:{all") + len("options:") :]
country_js = country_js[: country_js.index("},k={default")] country_js = country_js[: country_js.index("},k={default")]
country_tags = js_obj_str_to_python(country_js) country_tags = js_obj_str_to_python(country_js)
for k, v in country_tags.items(): for k, v in country_tags.items():
if k == 'all': if k == "all":
engine_traits.all_locale = 'all' engine_traits.all_locale = "all"
continue continue
country_tag = v['value'] country_tag = v["value"]
# add official languages of the country .. # add official languages of the country ..
for lang_tag in babel.languages.get_official_languages(country_tag, de_facto=True): for lang_tag in babel.languages.get_official_languages(country_tag, de_facto=True):
lang_tag = lang_map.get(lang_tag, lang_tag) lang_tag = lang_map.get(lang_tag, lang_tag)
sxng_tag = region_tag(babel.Locale.parse('%s_%s' % (lang_tag, country_tag.upper()))) sxng_tag = region_tag(babel.Locale.parse("%s_%s" % (lang_tag, country_tag.upper())))
# print("%-20s: %s <-- %s" % (v['label'], country_tag, sxng_tag)) # print("%-20s: %s <-- %s" % (v["label"], country_tag, sxng_tag))
conflict = engine_traits.regions.get(sxng_tag) conflict = engine_traits.regions.get(sxng_tag)
if conflict: if conflict:

9
searx/search/processors/online.py
View File

@ -155,8 +155,15 @@ class OnlineProcessor(EngineProcessor):
search_query.locale.language, search_query.locale.language,
) )
headers["Accept-Language"] = ac_lang headers["Accept-Language"] = ac_lang
self.logger.debug("HTTP Accept-Language: %s", headers.get("Accept-Language", "")) self.logger.debug("HTTP Accept-Language: %s", headers.get("Accept-Language", ""))
# https://developer.mozilla.org/en-US/docs/Glossary/Fetch_metadata_request_header
headers["Sec-Fetch-Dest"] = "empty"
headers["Sec-Fetch-Mode"] = "cors"
headers["Sec-Fetch-Site"] = "same-origin"
headers["Sec-Fetch-User"] = "?1"
headers["Sec-GPC"] = "1"
return params return params
def _send_http_request(self, params: OnlineParams): def _send_http_request(self, params: OnlineParams):