This switches the param used for the "country" config setting from "cr"
(which only filters results by the country the result is hosted in) to
"gl" (which overrides server/hosting location and produces results that
are more accurate for the user's current country).
Before this change, the country config setting was (imo) pretty useless.
Allowing a user to override an instance's hosting location with their
preferred country though is way more useful, especially for public
instances that are hosted in a different country than the user.
Closes#544
Previously the load/save/apply buttons in the config menu were hidden
below all available config options and required the user to scroll to
the bottom to save changes. This made for bad ux, since for new users,
it isn't immediately apparent that selecting a new dropdown value, for
instance, doesn't instantly save the new setting. The new layout should
make it more clear that hitting "Apply" is required to save config
changes.
This expands on the current testing suite a bit by introducing a new
workflow for testing functionality within the docker container. It runs
the same test suite as the regular "test" workflow, but also performs a
health check after running the app for 10 seconds to ensure
functionality.
The buildx workflow now waits for the docker test script to finish
successfully, rather than the regular test workflow. This will hopefully
avoid situations where new images are pushed with issues that aren't
detected in regular testing of the app.
Flask's `request.url` uses `http` as the protocol, which breaks
instances that enforce `https`, since the session redirect relies on
`request.url` for the follow-through URL.
This introduces a new method for determining the correct URL to use for
these redirects by automatically replacing the protocol with `https` if
the `HTTPS_ONLY` env var is set for that instance.
Fixes#538Fixes#545
HTTPS upgrades should be handled outside of Whoogle, since Flask often
doesn't detect the right protocol when being used behind a reverse proxy
such as Nginx.
This introduces a new approach to handling user sessions, which should
allow for users to set more reliable config settings on public instances.
Previously, when a user with cookies disabled would update their config,
this would modify the app's default config file, which would in turn
cause new users to inherit these settings when visiting the app for the
first time and cause users to inherit these settings when their current
session cookie expired (which was after 30 days by default I believe).
There was also some half-baked logic for determining on the backend
whether or not a user had cookies disabled, which lead to some issues
with out of control session file creation by Flask.
Now, when a user visits the site, their initial request is forwarded to
a session/<session id> endpoint, and during that subsequent request
their current session id is matched against the one found in the url. If
the ids match, the user has cookies enabled. If not, their original
request is modified with a 'cookies_disabled' query param that tells
Flask not to bother trying to set up a new session for that user, and
instead just use the app's fallback Fernet key for encryption and the
default config.
Since attempting to create a session for a user with cookies disabled
creates a new session file, there is now also a clean-up routine included
in the new session decorator, which will remove all sessions that don't
include a valid key in the dict. NOTE!!! This means that current user
sessions on public instances will be cleared once this update is merged
in. In the long run that's a good thing though, since this will allow session
mgmt to be a lot more reliable overall for users regardless of their cookie
preference.
Individual user sessions still use a unique Fernet key for encrypting queries,
but users with cookies disabled will use the default app key for encryption
and decryption.
Sessions are also now (semi)permanent and have a lifetime of 1 year.
Validation of the Tor connection occasionally fails with a
ConnectionError from requests, which was previously uncaught. This is
now handled appropriately (error message shown and connection dropped).
Fixes#532
This checks the latest released version of Whoogle against
the current app version, and shows an "update available"
message if the current version num < latest release num.
Closes#305
The config menu has gotten out of control recently, but rather than
reducing functionality, I'm just going to set a max height for the div
and allow scrolling within the menu.
Ultimately though this indicates that the app is getting a bit too
complicated (imo). Striking a balance between customization and
minimalism is less of a priority for me nowadays though, hence why I'm
willing to let it slide for now. At some point, maybe when there are
more contributors, it could be nice to refactor this in some way so that
it isn't overwhelming to new users who are looking to customize their
instance (that's just me speculating btw, I haven't actually heard from
anyone who thinks there are too many options in that menu).
Using `format` for formatting bang queries caused a KeyError for some
searches, such as !hd (HUDOC). In that example, the URL returned in the
bangs json was `http://...#{%22fulltext%22:[%22{}%22]...`, where
standard formatting would not work due to the misidentification of
"fulltext" as a formatting key.
The logic has been updated to just replace the first occurence of "{}"
in the URL returned by the bangs dict.
Fixes#513
Due to how the response is now reformed into a new bsoup object when
bolding search query terms, creating an ip card for "my ip" searches
threw an error due to how the new bsoup object was initialized for the
"my ip" card. This passes the response in as a string instead.
Fixes#504
DDG style bang searches can now have the bang (!) at the end of
the search (i.e. "bologna w!" will now redirect to wikipedia just like
"bologna !w" would)
Since the request class is loaded prior to values being read from the
user's dotenv, the WHOOGLE_RESULT_PER_PAGE var wasn't being used for
searches.
This moves the definition of the base search url to be intialized in the
request class to address this issue.
Fixes#497
variables.css doesn't need to be loaded by any template, since
WHOOGLE_CONFIG_STYLE loads those values by default when not set
explicitly. Loading the stylesheet caused the logo colors to be
persistent unless set individually.
Sorry @gripped for sneaking all of this unnecessary color in...
Fixes#492
This modifies the search result page by bold-ing all appearances
of any word in the original query. If portions of the query are in
quotes (i.e. "ice cream"), only exact matches of the sequence of
words will be made bold.
Co-authored-by: Ben Busby <noreply+git@benbusby.com>
Activating minimal mode should also remove all collapsed sections, if
any are found.
WHOOGLE_MINIMAL now documented in readme and app.json (for heroku).
I've gotten a bit bored of the current light/dark themes, so I'm
switching the default theme over to the Doppelganger theme, which is a
better template/jumping off point for users to use when creating custom
themes since it also provides examples for coloring each of the Whoogle
logo letters.
The levelup.gitconnected.com site is a Medium site that can also be
replaced with scribe.rip whenever privacy respecting site alternatives
are enabled in the config.
Also modified how link descriptions are updated when that config is
enabled (before it was missing replacements on quite a few
descriptions).
This introduces a new UI element for displaying the client IP
address when a search for "my ip" is used.
Note that this does not show the IP address seen by Google
if Whoogle is deployed remotely. It uses `request.remote_addr`
to display the client IP address in the UI, not the actual address
of the server (which is what Google sees in requests sent from
remote Whoogle instances).
scribe.rip is a privacy respecting front end for medium.com. This
feature allows medium.com results to be replaced with scribe.rip links,
and works for both regular medium.com domains as well as user specific
subdomains (i.e. user.medium.com).
[scribe.rip website](https://scribe.rip)
[scribe.rip source code](https://git.sr.ht/~edwardloveall/scribe)
Co-authored-by: Ben Busby <noreply+git@benbusby.com>
Used in header templates for navigating back to the home page when
behind a reverse proxy config where the app is running from a subpath of
a domain (i.e. "https://something/whoogle/")
Fixes#403
There are a few conventional choices but this one should be friendly
and generally accepted by local reader.
Previous version is still comprehensible but lesser users (perhaps
used in Japanese documents) and may give local users a pause.
Restricting form-action to 'self' in the content security policy
prevented Chrome (and likely other browsers) from using !bangs on the
home page.
Fixes#408
Previously if a result element marked for collapsing didn't have a valid
"parent" element, the collapsing was skipped altogether. This loops
through child elements until a valid parent is found (or if one isn't
found, the element will not be collapsed).
On app init, short hashes are generated from file checksums to use for
cache busting. These hashes are added into the full file name and used
to symlink to the actual file contents. These symlinks are loaded in the
jinja templates for each page, and can tell the browser to load a new
file if the hash changes.
This is only in place for css and js files, but can be extended in the
future for other file types if needed.
Introduces a new config element and environment variable
(WHOOGLE_CONFIG_THEME) for setting the theme of the app. Rather than
just having either light or dark, this allows a user to have their
instance use their current system light/dark preference to determine the
theme to use.
As a result, the dark mode setting (and WHOOGLE_CONFIG_DARK) have been
deprecated, but will still work as expected until a system theme has
been chosen.
Sections such as "People also asked" and "related searches" typically
take up a lot of room on the results page, and don't always have the
most useful information. This checks for result elements with more than
7 child divs, extracts the section title, and wraps all elements in a
"details" element that can be expanded/collapsed by the user.
Note that this functionality existed previously (albeit not implemented
as well), but due to changes in how Google returns searches (switching
from using <h2> elements for section headers to <span> or <div>
elements), the approach to collapsing these sections needed to be
updated.
* Add support for Lingva translations in results
Searches that contain the word "translate" and are normal search queries
(i.e. not news/images/video/etc) now create an iframe to a Lingva url to
translate the user's search using their configured search language.
The Lingva url can be configured using the WHOOGLE_ALT_TL env var, or
will fall back to the official Lingva instance url (lingva.ml).
For more info, visit https://github.com/TheDavidDelta/lingva-translate
* Add basic test for lingva results
* Allow user specified lingva instances through csp frame-src
* Fix pep8 issue
A recent issue brought up a good point about how the latest changes to
setting default language to english break functionality for bilingual
users. The change was likely not the best solution for users who were
being affected by IP geolocation on their instances -- the right
solution for that would be to configure the interface/search language to
their preference instead.
The requests library requires both 'http' and 'https' values in any
included proxy dict, and whoogle was previously copying the http proxy
to https for simplicity. The assumption was that if the underlying
request wasn't able to connect via https, it would default to http
(otherwise why have the requirement to specify both?)
This led to connectivity issues for users with http only proxies as of
the latest urllib and requests package versions, which are a lot more
strict with connections over https. With the latest versions, if an
https connection cannot be made, the library returns an error.
As a result, the new proxy dict must look something like this for plain
http proxies:
{'http': 'http://domain.tld:port', 'https': 'http://domain.tld:port'}
where both http and https are identical, but both are still required.
Since the interface language defaults to IP geolocation by google, the
default language is now set to english. Still not sure if this is the
best solution, but at least temporarily should clear up some confusion
for users with instances deployed in countries outside of their own.
Also performed some minor cleanup:
- Updated name of strip_blocked_sites to clean_query
- Added clean_query to list of jinja template functions
- Ensured site block list doesn't contain duplicate filters
Occasionally the search results will contain links with arguments such
as 'dq', which was being erroneously used in attempts to extract the 'q'
element from query strings. This enforces that only links with '?q=' or
'&q=' (elements with a standalone 'q' arg) will have the element
extracted.
I also refactored the naming of this element once extracted to be just
'q'. Although this seems counterintuitive, it makes a little more sense
since this element is the one we're extracting. It's a vague url arg
name, but it is what it is.
Bump version to 0.5.2 for hotfix release
The new site filter breaks links to Maps results, so filter.py needed
to be updated to handle these links as a unique case. A new method was
introduced to easily remove any "-site:..." filters from the query,
which is now also used to format queries in the header template rather
than manually removing the blocked site list within the template itself.
Bumps version to 0.5.1 for releasing the bugfix
Fixes#329
* Replace hardcoded strings using translation json file
This introduces a new "translations.json" file under app/static/settings
that is loaded on app init and uses the user config value for interface
language to determine the appropriate strings to use in Whoogle-specific
elements of the UI (primarily only on the home page).
* Verify interface lang can be used for localization
Check the configured interface language against the available
localization dict before attempting to use, otherwise fall back to
english.
Also expanded language names in the languages json file.
* Add test for validating translation language keys
Also adds Spanish translation to json (the only non-English language I
can add and reasonably validate on my own).
* Validate all translations against original keyset, update readme
Readme has been updated to include basic contributing guidelines for
both code and translations.
* add view image option
* prevent whoogle links from opening in a new tab.
* remove view image template on mobile requests
* change loop values to be more robust to the number of images
* Update app/templates/imageresults.html
* fix "Basically the .cvifge class needs width: 100%; in order to expand the search input to fit the form width."
* Update app/templates/imageresults.html
* remove hardcoded string from template
* Add view image config var to app.json
* Add view image config var to whoogle.env
Co-authored-by: jacr13 <ramos.joao@protonmail.com>
Co-authored-by: Ben Busby <benbusby@protonmail.com>
The wget method seemed to have a possible issue with creating endless
index.html copies (despite being specified to output to console only),
so this has been updated to use curl instead.
Also uses new non-authenticated "healthz" route to perform the
healthcheck.
Fix#316Fix#313
The previous method of removing all site filters from the search query
removed the last letter of the search. This only applies the substring
filter if any site filters are present in the query.
Fixes#306
* Block websites in search results via user config
Adds a new config field "Block" to specify a comma separated list of
websites to block in search results. This is applied for all searches.
* Add test for blocking sites from search results
* Document WHOOGLE_CONFIG_BLOCK usage
* Strip '-site:' filters from query in header template
The 'behind the scenes' site filter applied for blocked sites was
appearing in the query field when navigating between search categories
(all -> images -> news, etc). This prevents the filter from appearing in
all except "images", since the image category uses a separate header.
This should eventually be addressed when the image page can begin using
the standard whoogle header, but until then, the filter will still
appear for image searches.
* Add option to disable changing of configuration
Introduces a test to ensure the correct response code is found when
attempting to update the config when disabled, and ensure default config
is unchanged when posting a new config dict.
Attempting to update the config using the API when disabled now returns
a 403 code + redirect.
Co-authored-by: Ben Busby <benbusby@protonmail.com>
The search language is now set using the WHOOGLE_CONFIG_SEARCH_LANGUAGE
environment variable. Interface language is still set using
WHOOGLE_CONFIG_LANGUAGE.
Fixes#260
Enforces 0 margin for the search input form on the result page, which
removes the weird gap that is seen by default.
Also made minor changes to the border styling. Desktop searches now have
a single bottom border in dark mode rather than an all around border,
and the border around the mobile search result input was removed
entirely.
This was unfortunately a bit more complex than just adding an HTML reset
button, since reset buttons only "reset" input content to its original
value rather than clearing it. This doesn't work for Whoogle's needs,
since inputs on search result pages are auto populated with the search
content as their default value.
A reset button was introduced anyways, but is controlled by a few lines
of javascript to allow completely clearing the search input. The button
will only appear on mobile searches.
At the moment, it isn't particularly pretty, but is functional. It uses
just a plain "x" character and is always visible on mobile search result
pages. This leaves plenty of room for improvement moving forward.
Fixes#291
The recent change to cast bool config vars as ints to handle a '0' or
'1' value was shortsighted, since it doesn't allow for instances where
the variable is set to an empty value (or '' or any invalid/non-int
value).
This introduces a read_config_bool method for reading values that should
be a '0' or '1', but will default to False if not a digit (otherwise the
value will be cast as bool(int(value)) if "value" is a digit str).
Fixes#288
Config boolean environment variables need to be cast to ints, since
they are set or unset using 0 and 1. Previously they were interpreted as
(pseudocode) read_var(name, default=False), which meant that setting
CONFIG_VAR=0 would enable that variable since Python reads environment
variables as strings, and '0' is truthy. This updates the previous logic
to (still pseudocode) int(read_var(name, default='0')).
Fixes#279
Both light and dark themes have been updated to remove the leftover
hardcoded values (mostly related to the search suggestion styling).
See discussion in #247.
The logging from imported modules (stem, in particular) has caused quite
a few users to assume there are errors where there aren't any. The logs
from stem also aren't helpful, as everything in the library works as
expected despite the implication from the logs that it is not working.
Randomizing the "Mozilla" portion of the user agent changed the
character encoding to GB2312. Setting it to plain "Mozilla" enforces
UTF-8 encoding.
Bump to version 0.4.1 for release of bug fix
Fixes#267
This moves away from the previous (messy) approach of using two separate
keys for decrypting text and element URLs separately and regenerating
them for new searches. The current implementation of sessions is not very
reliable, which lead to keys being regenerated too soon, which would
break page navigation. Until that can be addressed, the single
key per session approach should work a lot better.
Fixes#250Fixes#90
The previous implementation of the is_heroku check in
search.needs_https() was implemented to only match URLs ending in
'.herokuapp.com', and skipped upgrading to HTTPS for other endpoints.
This introduces a set of environment variables that can be used for
defining initial config state, to expedite the process of
destroying/relaunching instances quickly with the same settings every
time.
Closes#228Closes#195
This allows the user to enable their preferred settings in a variety of
ways, depending on their deployment preference. Values added to
whoogle.env can be enabled using WHOOGLE_DOTENV=1, in which case all
values in the env var file will overwrite defaults or user provided
settings.
Co-authored-by: Ben Busby <benbusby@protonmail.com>
Eventually this should be part of a separate mypy ci build, but right
now it's just a general guideline. Future commits and PRs should be
validated for static typing wherever possible.
For reference, the testing commands used for this commit were:
mypy --ignore-missing-imports --pretty --disallow-untyped-calls app/
mypy --ignore-missing-imports --pretty --disallow-untyped-calls test/
* Add custom CSS field to config
This allows users to set/customize an instance's theme and appearance to
their liking. The config CSS field is prepopulated with all default CSS
variable values to allow quick editing.
Note that this can be somewhat of a "footgun" if someone updates the
CSS to hide all fields/search/etc. Should probably add some sort of
bandaid "admin" feature for public instances to employ until the whole
cookie/session issue is investigated further.
* Symlink all app static files to test dir
* Refactor app/misc/*.json -> app/static/settings/*.json
The country/language json files are used for user config settings, so
the "misc" name didn't really make sense. Also moved these to the static
folder to make testing easier.
* Fix light theme variables in dark theme css
* Minor style tweaking
The app/utils/*_utils weren't named very well, and all have been updated
to have more accurate names.
Function and class documention for the utils have been updated as well,
as part of the effort to improve overall documentation for the project.
Introduces a new content security policy header for responses to all
requests to reduce the possibility of ip leaks to outside connections.
By default blocks all inline scripts, and only allows content loaded
from Whoogle.
Refactors a few small inline scripting cases in the project to their own
individual scripts.
Requiring authentication for accessing the opensearch template prevents
the browser from accessing the file when adding as a default search
engine. This removes the authentication requirement from the opensearch
route, which should never provide any sensitive information anyways.
Bang operator can now be placed anywhere in the query, to allow for peak
efficiency in stream of consciousness querying (i.e. `big !reddit
chungus` will search reddit for big chungus`).
Fixes#196
* Adds the ability to redirect reddit.com to libredd.it using the existing
"site alts" config setting.
This adds the WHOOGLE_ALT_RD environment variable for optionally
redirecting reddit links to libreddit
(https://github.com/spikecodes/libreddit).
* Include libreddit in home page site alt note
Pip installs of whoogle search were missing access to the misc/ folder,
which previously contained the language and country json files. These
have been moved to app/misc, and the previous root level misc/ was
renamed to config/ (since it now only contains the tor config files).
Bump to 0.3.1.
Heroku instances were using the base http url when formatting the
opensearch.xml template. This adds a new routing utility, "needs_https",
which can be used for determining if the url in question needs
upgrading.
With javascript disabled, searches could not be submitted on the results
page using the "Enter" key. Adding a hidden submit button to the header
template resolves this issue.
The lxml dependency in the project was fairly unnecessary, and made the
initial build time for the project considerably slower. This replaces
all instances of lxml with either the default html parser (for bs4
constructors) or the built in xml.etree package (for search suggestion
parsing).
Introduces a new javascript "utils" file, which includes a check for
matching the query against a set of tracking number regexes on page
load. If a match is found, the script prepends a link to the
(presumably) appropriate tracking page.
Referenced in #98
Moves the language and country dicts from the config model to json files
that are loaded during app init and stored in the app config dict. This
substantially improves the readability of the config model and allows
for much more sensible loading of the language/country options.
This adds a step in the filter process to wrap the "people also ask"
section in a <details> element, which automatically collapses the
contents of the section. Clicking/tapping the details element expands
the view as normal.
See #113
The BeautifulSoup constructur in gen_nojs needed to explicitly set
features='lxml' to silence a warning from the library.
Also temporarily disabled the site alts test since the results are too
unreliable. This should be moved to a unit test instead.
Ben Busby
Ilya Prokopenko
Joao A. Candido Ramos
gnuhead-chieb
Fabian Schilling
Vansh Comar
gripped
DUO Labs
Yadomin
Albony Cal
rn83
BorislavGeorgiev
BlissOWL
Kang-min Liu
alefvanoon
davidfrickert
Trottel
Darkempire
Laurent le Beau-Martin
සයුරි | Sayuri
Ben Busby
deluxghost
Aikatsui
Ricardo
bruvv
Myzel394
Federico Torrielli
Harsh Barsaiyan
Angel Mario
gripped
Shimul
Tomasz Borychowski
Roman Štefko
Ziga Zajc