## What does this PR do?
This PR adds search operator plugin to searx. By default it is disabled because it
removes results from your result set. Thus, you might end up with an empty result page with
the additional filtering.
## Why is this change important?
With all of its shortcomings, still is a nifty plugin.
## How to test this PR locally?
```
batman -site:imdb.com
```
Co-authored-by: DiamondDemon669 <62653580+DiamondDemon669@users.noreply.github.com>
Adds an option to the oscar theme that puts an archive.today link next to the web.archive.org cache link. In preferences change 'Show archive.today links' to 'On'
* [fix] make autofocus configurable (#1984)
"Results page: having text cursor still in search field is annoying #1984"
Allows autofocus on the results page to be configured either in the UI
or in the searx server config.
* fix commented code
Co-authored-by: Ben Collerson <benc@benon.com>
* [fix] google engine: results XPath
* [fix] google & youtube - set EU consent cookie
This change the previous bypass method for Google consent using
``ucbcb=1`` (6face215b8) to accept the consent using ``CONSENT=YES+``.
The youtube_noapi and google have a similar API, at least for the consent[1].
Get CONSENT cookie from google reguest::
curl -i "https://www.google.com/search?q=time&tbm=isch" \
-A "Mozilla/5.0 (X11; Linux i686; rv:102.0) Gecko/20100101 Firefox/102.0" \
| grep -i consent
...
location: https://consent.google.com/m?continue=https://www.google.com/search?q%3Dtime%26tbm%3Disch&gl=DE&m=0&pc=irp&uxe=eomtm&hl=en-US&src=1
set-cookie: CONSENT=PENDING+936; expires=Wed, 24-Jul-2024 11:26:20 GMT; path=/; domain=.google.com; Secure
...
PENDING & YES [2]:
Google change the way for consent about YouTube cookies agreement in EU
countries. Instead of showing a popup in the website, YouTube redirects the
user to a new webpage at consent.youtube.com domain ... Fix for this is to
put a cookie CONSENT with YES+ value for every YouTube request
[1] https://github.com/iv-org/invidious/pull/2207
[2] https://github.com/TeamNewPipe/NewPipeExtractor/issues/592
Closes: https://github.com/searxng/searxng/issues/1432
* [fix] sjp engine - convert enginename to a latin1 compliance name
The engine name is not only a *name* its also a identifier that is used in
logs, HTTP headers and more. Unicode characters in the name of an engine could
cause various issues.
Closes: https://github.com/searxng/searxng/issues/1544
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
* [fix] engine tineye: handle 422 response of not supported img format
Closes: https://github.com/searxng/searxng/issues/1449
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
* bypass google consent with ucbcb=1
* [mod] Adds Lingva translate engine
Add the lingva engine (which grabs data from google translate). Results from
Lingva are added to the infobox results.
* openstreetmap engine: return the localized named.
For example: display "Tokyo" instead of "東京都" when the language is English.
* [fix] engines/openstreetmap.py typo: user_langage --> user_language
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
* Wikidata engine: ignore dummy entities
* Wikidata engine: minor change of the SPARQL request
The engine can be slow especially when the query won't return any answer.
See https://www.mediawiki.org/wiki/Wikidata_Query_Service/User_Manual/MWAPI#Find_articles_in_Wikipedia_speaking_about_cheese_and_see_which_Wikibase_items_they_correspond_to
Co-authored-by: Léon Tiekötter <leon@tiekoetter.com>
Co-authored-by: Emilien Devos <contact@emiliendevos.be>
Co-authored-by: Markus Heiser <markus.heiser@darmarit.de>
Co-authored-by: Emilien Devos <github@emiliendevos.be>
Co-authored-by: ta <alt3753.7@gmail.com>
Co-authored-by: Alexandre Flament <alex@al-f.net>
## What does this PR do?
This PR prepares for removing `httpx`, and reverts back to `requests`.
## Why is this change important?
`httpx` hasn't proven itself to be faster or better than `requests`. On the other
hand it has caused issues on Windows.
=============================================
Please update your environment to use requests instead of httpx.
=============================================
* [fix] Rename ccengine engine to openverse
The CC engine was merged with WordPress and renamed to Openverse
Source: https://wordpress.org/news/2021/05/welcome-to-openverse/
* [fix] ccengine engine - avoid unwanted redirects
api.openverse.engineering is a little picky and wants to have a trailing slash
in the path:
/v1/images? -->/ v1/images/?
otherwise it redirects, here is the debug log:
DEBUG searx.network.openverse : HTTP Request: GET https://api.openverse.engineering/v1/images?&page=1&page_size=20&format=json&q=foo "HTTP/2 301 Moved Permanently" (text/html; charset=utf-8)
DEBUG searx.network.openverse : HTTP Request: GET https://api.openverse.engineering/v1/images/?&page=1&page_size=20&format=json&q=foo "HTTP/2 200 OK" (application/json)
WARNING searx.engines.openverse : ErrorContext('searx/search/processors/online.py', 105, 'count_error(', None, '1 redirects, maximum: 0', ('200', 'OK', 'api.openverse.engineering')) True
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
* [fix] FutureWarning from lxml
Just in case if content is None, the original code will skip extract_text(), and
just append the None value to 'content'. So just add allow_none=True, and this
will return None without raising a ValueError in extract_text().
* [enh] Add pagination to Brave
Also added ```&spellcheck=1``` because now it is disabled by default, not returning any ```suggestion_xpath```.
Co-authored-by: Léon Tiekötter <leon@tiekoetter.com>
Co-authored-by: Markus Heiser <markus.heiser@darmarit.de>
Co-authored-by: capric98 <42015599+capric98@users.noreply.github.com>
Co-authored-by: Allen <64094914+allendema@users.noreply.github.com>
## What does this PR do?
This should fix#3164.
The problem is that `httpx` keeps making breaking changes to their library, so we just have to adjust the code a little bit to make it work with the new version of the library.
## Related issues
Closes #3164
* [mod] /image_proxy: don't decompress images
* [fix] image_proxy: always close the httpx respone
previously, when the content type was not an image and some other error,
the httpx response was not closed
* [mod] /image_proxy: use HTTP/1 instead of HTTP/2
httpx: HTTP/2 is slow when a lot data is downloaded.
https://github.com/dalf/pyhttp-benchmark
also, the usage of HTTP/1 decreases the load average
* [mod] searx.utils.dict_subset: rewrite with comprehension
Co-authored-by: Alexandre Flament <alex@al-f.net>
* [enh] Add Tineye reverse image search
Other optional parametesr:
"&sort=crawl_date" can be appended to search_string to sort results by date.
"&domain=example.org" can be implemented to search_string to get results from just one domain.
Public instances could get relatively fast timed-out for 3600s.
* [enh] Add TIneye to settings.yml
Check if that's the right shortcut.
* [mod] Fix checks
* [mod] Try to fix checks
* [mod] Use Four spaces for indentation
And set paging back to True
Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com>
Noémi Ványi
dependabot[bot]
searx-bot
Noémi Ványi
Ben Collerson
james-still
Ben Collerson
Adam Tauber
liimee
Eric Zhang
Markus Heiser
Marc Abonce Seguin
israelyago
iko
Alexandre Flament
Allen