The fundamental goal in this refactor is to make it easier to disable the
Listings feature set. By extracting a routes file (via the `draw` method [see
implementation details][1]), we can more easily disable those routes _en
masse_. This "crease in the code" could mean that we wouldn't need to update
the controller action logic. An update of that logic might mean going into
each controller action related to listing and adding a conditional for render
404 when listing is disabled.
Now, on to why this should work (not I did not open the application and click
around, but instead went with a "let the framework help me" mindset).
I believe that this should work; though there's a caveat†. Here's why I
believe this will work for extracting the listing routes:
Before making changes I ran the following:
```sh
$ rails routes | sort > original-routes.txt
```
Let's break that down:
- `rails routes` :: renders the routing table for Rails; this is how
Rails converts routes to the magic of stuff like `article_path(:id)`
- `| sort` :: to sort the results
- `> original-routes.txt` :: to write the results into a file.
Then I made the changes and ran the following:
```sh
$ rails routes | sort > updated-listing-routes.txt
```
And finally, I ran:
```sh
$ diff updated-routes.txt original-routes.txt
```
The results were as follows:
```sh
500c500
< api_organization_listings GET (/locale/:locale)/api/organizations/:organization_username/listings(.:format) api/v0/organizations#listings {:locale=>nil, :format=>"json", :to=>"organizations#listings"}
---
> api_organization_listings GET (/locale/:locale)/api/organizations/:organization_username/listings(.:format) api/v0/organizations#listings {:locale=>nil, :format=>"json"}
```
Note the only difference is that the above two lines, one has
`:to=>"organizations#listings"` and the other does not. However, the
`./spec/requests/api/v0/organizations_spec.rb` exercises the route; so I
assume a hiccup/foible in the route reporting.
There was no difference between the before and after files.
† Caveat: if for some reason the order of the routes mattered, this will
have messed that up. I don't think that is the case, but we shall see.
Related to forem/rfcs#291 and #16335
* Added tag search to nav menu
* Added tag search
* Improved tags search results view
* Removed commented lines from the controller
* Fix specs for Search::Tag
* Prepare for tags search pagination
* Fixed Search::Tag specs
* styling
Co-authored-by: Paweł Ludwiczak <ludwiczakpawel@gmail.com>
* Convert symbol hash keys to strings when calling .perform_async
Fixes a warning from Sidekiq 6.4.0+ about perform_async arguments
which are not equal when passed to perform (`JSON.parse(JSON.dump(arg))`
should equal arg).
This is a safety measure to prevent passing objects (like classes, or
model instances) rather than their representations (like a class name,
or a model's attributes hash).
This warning will be an error in sidekiq 7
* Turn warning into an error in non-production environments
* Use string keys for reaction notification and article fetched
Missed these two on the first pass
* Update example argument hashes for #enqueues_on_correct_queue
Since this calls perform_async under the hood we need to pass json
safe hashes in the test cases as well.
https://github.com/forem/forem/blob/main/spec/workers/shared_examples/enqueues_on_correct_queue.rb
* Convert keys from FollowData#to_h to string before perform_async
I'm not sure enough where else (outside of notification) to_h is being
called, so I'm converting here when building args, rather than in
FollowData#to_h, which might be my next step.
* Let FollowData#to_h return a hash with string keys
Update spec to use string keys as well.
* Make to_h return string keys for ReactionData
Like FollowData, the #to_h method is only used to call
notifications (this is used to enqueue sidekiq jobs).
* Remove a key that was in the hash
Since reaction_data calls to_h, it gets string and not symbol,
keys. Call Hash#except with a key that was actually there.
* Initial work
* Implement strategy injection into WeightedQueryStrategy
* Modify field_test config and adding variants
* Change orginal to constant and make some other adjustments
* Fix hardcoded test values
* Proposing a new feed experiment
This commit involves tweaking a few subtle aspects of the challenger:
1) Remove the sort by published date on the relevance score. The
articles will now be listed in the order of perceived relevance.
2) Disable a few levers that did little. There really aren't any
`articles.featured = true` articles in the database.
3) Gently increase the weight of each comment in a linear manner.
4) Give a little more weight to posts that don't have followed tags.
In addition, it involves renaming the conversions to better convey their
implementation. This renaming helps create more descriptive labels for
the test results at `/admin/abtests`.
This commit also adds logic to repurpose the AbExperiment feed_strategy
logic; I envision adjusting the weighted feed strategy levers with some
frequency.
Per discussions, I'm also disabling the "not logged in" feed testing.
We'll use the LargeForemExperimental for this.
* Flipping attribute name to positive
Mentally "not disabled" is harder to parse than "enabled". This change
helps with setting an optimistic attribute.
* Bump for travis
* Bump for travis
* Removing file committed by mistake
* Extracting method
* test: whitespace unicode characters cannot be used as titles or tags
* feat: add localized error message
* refactor: use localized error message
* fix: whitespace unicode characters cannot be used as titles or tags
* chore: fix locale after merge
* refactor: fix indentation
* Fix spelling of prohibited in method names
Additionally, rubocop removed a redundant user_id validation since
Article belongs to user.
* Fix spelling
Missed one method call on the same line (fixed the first of two
instances needing changes)
* add failing test cases
The reorganization to remove let was due to a limit on nesting rspec
contexts (it inside context inside describe inside describe when
trying to use different titles in let blocks in contexts)
The first test was clarified (the "U+202D" string that looks like the
code for a unicode point is valid, but the character \u202d is
expected to be invalid
The remaining two tests are based on the feedback I'd given in the PR,
failing because we don't assert title is present after removing
unicode whitespace, and we don't actually set the title (gsub is
non-destructive, returning a value).
* Set title to sanitized title
contains_prohibitied_unicode_characters? was using == (which is not a
good match for strings to regexes), use =~ instead
Change invalid example characters in titles from \u202d (bidi override) to
\u200a (hair space)
Assert that the empty title is blank and can't be blank (even though
it contains no prohibitd characters after replacement)
* change the definition of prohibited characters
From the description, it looks like "unicode space characters" was the
desired rejection set. It was unclear what the existing regex was
matching on (I couldn't get the expected examples to match correctly).
Given the intent, I select "unicode space property, except the ascii
space character", which may _also_ be incorrect but passed the tests.
* Remove now-invalid expectation for presence of user id
Since this was redundant (belongs_to user), we no longer will validate
presence of user id, and we should not test for it.
* Only reject bidirectional text controls
The original issue pointed to the BIDI controls as problematic.
While they called out other "whitespace" characters, as we accept a
wider range of input languages, the need to handle non-printing
punctuation (for example a group separating space for some asian
languages).
It's possible a wider list of characters should be added - if that's
the case I suggest this regex and the sanitization be moved to a
standalone class, and each of the recommendations in
https://www.w3.org/TR/unicode-xml/#Charlist be checked for
applicability.
* Check for blank title after sanitizing disallowed characters
Since validate_title modifies the title, and doesn't set error
messages on the model itself, check for non-empty title _after_
potentially removing any disallowed characters.
* Remove invalid characters before validation
Remove the validate_title and contains_prohibited_characters methods,
and center all logic on the remove prohibitied unicode characters
method.
Remove unneeded and unused arguments (input string is always title,
replacement is always removal/empty string).
* handle case when validating null title
If the title's nil, we don't want to call match? since it will fail.
Title will sometimes be nil when validating.
Co-authored-by: Ben Halpern <bendhalpern@gmail.com>
Co-authored-by: Dan Uber <dan@forem.com>
We were previously disallowing empty user agents, but we weren't
blocking requests where the User-Agent header wasn't set at all. This
commit blocks those requests.
This commit also blocks Chrome 74. It's a 2.5-year-old version of an
evergreen browser that releases every 6 weeks, so this is clearly a bot
spoofing this header.
* Removing the stackbit integration
The feature didn't quite work and Stackbit no longer supports this
integration.
Yes there are a few places where the webhooks has a string of "stackbit"
but I'm hesitant to remove that, as they are the part of the Webhooks
tests.
Closes#15700
* Update lib/data_update_scripts/20211206222716_remove_stackbit_page.rb
Co-authored-by: Michael Kohl <citizen428@forem.com>
Co-authored-by: Michael Kohl <citizen428@forem.com>
