Typification of SearXNG
=======================
This patch introduces the typing of the results. The why and how is described
in the documentation, please generate the documentation ..
$ make docs.clean docs.live
and read the following articles in the "Developer documentation":
- result types --> http://0.0.0.0:8000/dev/result_types/index.html
The result types are available from the `searx.result_types` module. The
following have been implemented so far:
- base result type: `searx.result_type.Result`
--> http://0.0.0.0:8000/dev/result_types/base_result.html
- answer results
--> http://0.0.0.0:8000/dev/result_types/answer.html
including the type for translations (inspired by #3925). For all other
types (which still need to be set up in subsequent PRs), template documentation
has been created for the transition period.
Doc of the fields used in Templates
===================================
The template documentation is the basis for the typing and is the first complete
documentation of the results (needed for engine development). It is the
"working paper" (the plan) with which further typifications can be implemented
in subsequent PRs.
- https://github.com/searxng/searxng/issues/357
Answer Templates
================
With the new (sub) types for `Answer`, the templates for the answers have also
been revised, `Translation` are now displayed with collapsible entries (inspired
by #3925).
!en-de dog
Plugins & Answerer
==================
The implementation for `Plugin` and `Answer` has been revised, see
documentation:
- Plugin: http://0.0.0.0:8000/dev/plugins/index.html
- Answerer: http://0.0.0.0:8000/dev/answerers/index.html
With `AnswerStorage` and `AnswerStorage` to manage those items (in follow up
PRs, `ArticleStorage`, `InfoStorage` and .. will be implemented)
Autocomplete
============
The autocompletion had a bug where the results from `Answer` had not been shown
in the past. To test activate autocompletion and try search terms for which we
have answerers
- statistics: type `min 1 2 3` .. in the completion list you should find an
entry like `[de] min(1, 2, 3) = 1`
- random: type `random uuid` .. in the completion list, the first item is a
random UUID
Extended Types
==============
SearXNG extends e.g. the request and response types of flask and httpx, a module
has been set up for type extensions:
- Extended Types
--> http://0.0.0.0:8000/dev/extended_types.html
Unit-Tests
==========
The unit tests have been completely revised. In the previous implementation,
the runtime (the global variables such as `searx.settings`) was not initialized
before each test, so the runtime environment with which a test ran was always
determined by the tests that ran before it. This was also the reason why we
sometimes had to observe non-deterministic errors in the tests in the past:
- https://github.com/searxng/searxng/issues/2988 is one example for the Runtime
issues, with non-deterministic behavior ..
- https://github.com/searxng/searxng/pull/3650
- https://github.com/searxng/searxng/pull/3654
- https://github.com/searxng/searxng/pull/3642#issuecomment-2226884469
- https://github.com/searxng/searxng/pull/3746#issuecomment-2300965005
Why msgspec.Struct
==================
We have already discussed typing based on e.g. `TypeDict` or `dataclass` in the past:
- https://github.com/searxng/searxng/pull/1562/files
- https://gist.github.com/dalf/972eb05e7a9bee161487132a7de244d2
- https://github.com/searxng/searxng/pull/1412/files
- https://github.com/searxng/searxng/pull/1356
In my opinion, TypeDict is unsuitable because the objects are still dictionaries
and not instances of classes / the `dataclass` are classes but ...
The `msgspec.Struct` combine the advantages of typing, runtime behaviour and
also offer the option of (fast) serializing (incl. type check) the objects.
Currently not possible but conceivable with `msgspec`: Outsourcing the engines
into separate processes, what possibilities this opens up in the future is left
to the imagination!
Internally, we have already defined that it is desirable to decouple the
development of the engines from the development of the SearXNG core / The
serialization of the `Result` objects is a prerequisite for this.
HINT: The threads listed above were the template for this PR, even though the
implementation here is based on msgspec. They should also be an inspiration for
the following PRs of typification, as the models and implementations can provide
a good direction.
Why just one commit?
====================
I tried to create several (thematically separated) commits, but gave up at some
point ... there are too many things to tackle at once / The comprehensibility of
the commits would not be improved by a thematic separation. On the contrary, we
would have to make multiple changes at the same places and the goal of a change
would be vaguely recognizable in the fog of the commits.
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
The restriction of shell scripts to certain distributions is only required for
certain actions such as the installation of a SearXNG instance. The maintenance
scripts and build processes were previously also restricted to these specific
distributions. With this patch, the build processes (such as the build of
online documentation) can now also be executed on all Linux distributions.
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
Another trip into the hell of dependencies: docutils tends to put major changes
in minor patches: the executables have been renamed / e.g.
rst2html.py --> rts2html
so we have to use docutils at least from version 0.21.2, but this version of
docutils is only supported by myst-parser from version 3.0.1 on.
Additionally, docutils decided to drop python 3.8 in version 0.21 [1]
Further, linuxdoc needed an update to cope with docutils 0.21 [2]
[1] https://docutils.sourceforge.io/RELEASE-NOTES.html#release-0-21-2024-04-09
[2] https://github.com/return42/linuxdoc/pull/36
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
In the past, some files were tested with the standard profile, others with a
profile in which most of the messages were switched off ... some files were not
checked at all.
- ``PYLINT_SEARXNG_DISABLE_OPTION`` has been abolished
- the distinction ``# lint: pylint`` is no longer necessary
- the pylint tasks have been reduced from three to two
1. ./searx/engines -> lint engines with additional builtins
2. ./searx ./searxng_extra ./tests -> lint all other python files
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
Add a leading tag (in square brackets) about the scope/type to commit messages
from automated tasks (commits from CI).
dependantbot::
[upd] pypi: Bump .. from .. to ..
[upd] npm: Bump .. from .. to .. in /searx/static/themes/simple
Weblate translation updates::
[l10n] update translations from Weblate
updates of ./data::
[data] update searx.data ...
build commit of gh-pages::
[doc] build from commit ...
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
babel.Locale.parse loads more than 60MB in RAM. The only purpose is to get:
LOCALE_NAMES - searx.data.LOCALES["LOCALE_NAMES"]
RTL_LOCALES - searx.data.LOCALES["RTL_LOCALES"]
This commit calls babel.Locale.parse when the translations are update from
weblate and stored in::
searx/data/locales.json
This file can be build by::
./manage data.locales
By store these variables in searx.data when the translations are updated we save
round about 65MB (usually 4 worker = 260MB of RAM saved.
Suggested-by: https://github.com/searxng/searxng/discussions/2633#discussioncomment-8490494
Co-authored-by: Markus Heiser <markus.heiser@darmarit.de>
All the environments defined in ./utils/brand.env are generated on the fly, so
there is no longer a need to define the brand environment in this file and all
the workflows to handle this file.
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
This allow to read settings on the fly even without virtualenv. The ultimate
goal of the commit is to remove utils/brand.env from the git repository.
The code includes a tiny yaml parser that **should** be good enough. The code
read searx/settings.yml directly (and ignore the environment variables).
yq [1] is a more reliable alternative but this require to download a binary from
github which is not great.
[1] https://github.com/mikefarah/yq/#install
* Docker: add UWSGI_WORKERS and UWSGI_THREAD.
UWSGI_WORKERS specifies the number of process.
UWSGI_THREADS specifies the number of threads.
The Docker convention is to specify the whole configuration
through environment variables. While not done in SearXNG, these two
additional variables allows admins to skip uwsgi.ini
In additional, https://github.com/searxng/preview-environments starts Docker
without additional files through searxng-helm-chat.
Each instance consumes 1Go of RAM which is a lot especially when there are a
lot of instances / pull requests.
* [scripts] add environments UWSGI_WORKERS and UWSGI_THREADS
- UWSGI_WORKERS specifies the number of process.
- UWSGI_THREADS specifies the number of threads.
Templates for uwsgi scripts can be tested by::
UWSGI_WORKERS=8 UWSGI_THREADS=9 \
./utils/searxng.sh --cmd\
eval "echo \"$(cat utils/templates/etc/uwsgi/*/searxng.ini*)\""\
| grep "workers\|threads"
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
---------
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
Co-authored-by: Markus Heiser <markus.heiser@darmarit.de>
Caching files on the client side for more than a day can confuse the end user
when updating static files[1].
Depending on the way of providing a SearXNG instance via HTTP, there are several
ways to optimize the access to the /static files. However, since we don't know
what optimization an admin has provided for his static files, we should have
moderate settings in the defaults that run robustly in a wide variety of
installations.
In this sense, all caches on the client side should be cleared after one day at
the latest. So far the files were cached for one year on client side; as soon
as changes are made to the static files (with the option `static_use_hash:
true`) the old static files are kept for one year on the CLient side / which can
also be evaluated as unnecessary caching.
[1] https://github.com/searxng/searxng/discussions/2821
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
BTW force modularization of the ./mange script into sub modules:
- utils/lib_sxng_data.sh
- utils/lib_sxng_node.sh
- utils/lib_sxng_static.sh
- utils/lib_sxng_test.sh
- utils/lib_sxng_themes.sh
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
Partial reverse engineering of the Google engines including a improved language
and region handling based on the engine.traits_v1 data.
When ever possible the implementations of the Google engines try to make use of
the async REST APIs. The get_lang_info() has been generalized to a
get_google_info() function / especially the region handling has been improved by
adding the cr parameter.
searx/data/engine_traits.json
Add data type "traits_v1" generated by the fetch_traits() functions from:
- Google (WEB),
- Google images,
- Google news,
- Google scholar and
- Google videos
and remove data from obsolete data type "supported_languages".
A traits.custom type that maps region codes to *supported_domains* is fetched
from https://www.google.com/supported_domains
searx/autocomplete.py:
Reversed engineered autocomplete from Google WEB. Supports Google's languages and
subdomains. The old API suggestqueries.google.com/complete has been replaced
by the async REST API: https://{subdomain}/complete/search?{args}
searx/engines/google.py
Reverse engineering and extensive testing ..
- fetch_traits(): Fetch languages & regions from Google properties.
- always use the async REST API (formally known as 'use_mobile_ui')
- use *supported_domains* from traits
- improved the result list by fetching './/div[@data-content-feature]'
and parsing the type of the various *content features* --> thumbnails are
added
searx/engines/google_images.py
Reverse engineering and extensive testing ..
- fetch_traits(): Fetch languages & regions from Google properties.
- use *supported_domains* from traits
- if exists, freshness_date is added to the result
- issue 1864: result list has been improved a lot (due to the new cr parameter)
searx/engines/google_news.py
Reverse engineering and extensive testing ..
- fetch_traits(): Fetch languages & regions from Google properties.
*supported_domains* is not needed but a ceid list has been added.
- different region handling compared to Google WEB
- fixed for various languages & regions (due to the new ceid parameter) /
avoid CONSENT page
- Google News do no longer support time range
- result list has been fixed: XPath of pub_date and pub_origin
searx/engines/google_videos.py
- fetch_traits(): Fetch languages & regions from Google properties.
- use *supported_domains* from traits
- add paging support
- implement a async request ('asearch': 'arc' & 'async':
'use_ac:true,_fmt:html')
- simplified code (thanks to '_fmt:html' request)
- issue 1359: fixed xpath of video length data
searx/engines/google_scholar.py
- fetch_traits(): Fetch languages & regions from Google properties.
- use *supported_domains* from traits
- request(): include patents & citations
- response(): fixed CAPTCHA detection (Scholar has its own CATCHA manager)
- hardening XPath to iterate over results
- fixed XPath of pub_type (has been change from gs_ct1 to gs_cgt2 class)
- issue 1769 fixed: new request implementation is no longer incompatible
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
$ make nvm.install
INFO: install (update) NVM at /800GBPCIex4/share/SearXNG/.nvm
INFO: already cloned at: /800GBPCIex4/share/SearXNG/.nvm
|| Fetching origin
INFO: checkout v0.39.1
|| HEAD is now at 9600617 v0.39.1
make: *** [Makefile:96: nvm.install] Error 1
Without this fix we need to set VERBOSE environment to avoid the 'Error 1':
$ VERBOSE=0 make nvm.install
BTW: fix an issue if there are any leftovers in ${NVM_DIR} from previos
installations
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
There's already precedence for not using GNUism sha256sum longopts as
seen in searxng/utils/lib_go.sh so update lib.sh to not use them either.
A nice side effect is now the sha256sum usage doesn't care if you're
using BSD sha256sum or GNU sha256sum which makes this work under FreeBSD.
settings.yml:
* The default URL was unix:///usr/local/searxng-redis/run/redis.sock?db=0
* The default URL is now "false"
The default URL makes the log difficult to deal with:
if the admin didn't install a Redis instance, the logs record a false error.
It worked before because SearXNG initialized the Redis connection when the limiter started.
In this commit, SearXNG initializes Redis in searx/webapp.py
so various components can use Redis without taking care of the initialization step.
Since ./utils/searxng.sh is implemented, the old installation procedures from
filtron, morty and searx can be removed.
For users who want to upgrade, the procedures for removing old installations
have still been retained.
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
