Users who use encoded slugs on their sites sometimes run into 500 error when pasting a link to another topic in a post. The problem happens when generating a backward "reflection" link that would appear in a linked topic. Link URL restricted on the database level to 500 chars in length. At first glance, it should work since we have a restriction on topic title length.
But it doesn't work when a site uses encoded slugs, like here (take a look at the URL). The link to a topic, in this case, can be much longer than 500 characters.
By the way, an error happens only when generating a "reflection" link and doesn't happen with a direct link, we truncate that link. It works because, in this case, the original long link is still present in the post body and can be used for navigation. But we can't do the same for backward "reflection" links (without rewriting their implementation), the whole link must be saved to the database.
The simplest and cleanest solution will be just to remove the restriction on the database level. Abuse is impossible here since we are already protected by the restriction on topic title length. There aren’t performance benefits in using length-constrained columns in Postgres, in fact, length-constrained columns need a few extra CPU cycles to check the length when storing data.
Previously the code was very race condition prone leading to
odd failures in production
It was re-written in raw SQL to avoid conditions where rows
conflict on inserts
There is no clean way in ActiveRecord to do:
Insert, on conflict do nothing and return existing id.
This also increases test coverage, we were previously not testing
the code responsible for crawling external sites directly
* Introduced fab!, a helper that creates database state for a group
It's almost identical to let_it_be, except:
1. It creates a new object for each test by default,
2. You can disable it using PREFABRICATION=0
This change both speeds up specs (less strings to allocate) and helps catch
cases where methods in Discourse are mutating inputs.
Overall we will be migrating everything to use #frozen_string_literal: true
it will take a while, but this is the first and safest move in this direction
This reverts commit 993f847a2c.
There is an edge case where the link click redirect fails when the URL has trailing slash. Need to figure out a better fix for this.
This was an indentation mistake introduced in 44eba0b. Pretty understandable, considering we are indented 8 levels deep in this method. Will follow-up with a refactor to improve this.
Introduce new patterns for direct sql that are safe and fast.
MiniSql is not prone to memory bloat that can happen with direct PG usage.
It also has an extremely fast materializer and very a convenient API
- DB.exec(sql, *params) => runs sql returns row count
- DB.query(sql, *params) => runs sql returns usable objects (not a hash)
- DB.query_hash(sql, *params) => runs sql returns an array of hashes
- DB.query_single(sql, *params) => runs sql and returns a flat one dimensional array
- DB.build(sql) => returns a sql builder
See more at: https://github.com/discourse/mini_sql
Since rspec-rails 3, the default installation creates two helper files:
* `spec_helper.rb`
* `rails_helper.rb`
`spec_helper.rb` is intended as a way of running specs that do not
require Rails, whereas `rails_helper.rb` loads Rails (as Discourse's
current `spec_helper.rb` does).
For more information:
https://www.relishapp.com/rspec/rspec-rails/docs/upgrade#default-helper-files
In this commit, I've simply replaced all instances of `spec_helper` with
`rails_helper`, and renamed the original `spec_helper.rb`.
This brings the Discourse project closer to the standard usage of RSpec
in a Rails app.
At present, every spec relies on loading Rails, but there are likely
many that don't need to. In a future pull request, I hope to introduce a
separate, minimal `spec_helper.rb` which can be used in tests which
don't rely on Rails.
update rspec syntax to v3
change syntax to rspec v3
oops. fix typo
mailers classes with rspec3 syntax
helpers with rspec3 syntax
jobs with rspec3 syntax
serializers with rspec3 syntax
views with rspec3 syntax
support to rspec3 syntax
category spec with rspec3 syntax
FIX: history revision can now properly be hidden
FIX: PostRevision serializer is now entirely dynamic to properly handle
hidden revisions
FIX: default history modal to "side by side" view on mobile
FIX: properly hiden which revision has been hidden
UX: inline category/user/wiki/post_type changes with the revision
details
FEATURE: new '/posts/:post_id/revisions/latest' endpoint to retrieve
latest revision
UX: do not show the hide/show revision button on mobile (no room for
them)
UX: remove CSS transitions on the buttons in the history modal
FIX: PostRevisor now handles all the changes that might create new
revisions
FIX: PostRevision.ensure_consistency! was wrong due to off by 1
mistake...
refactored topic's callbacks for better readability
extracted 'PostRevisionGuardian'
When accessed over IPv6, the ip address of the user is a 128-bit number,
too big for PostgreSQL's bigint data type. Since PostgresSQL has the
built-in inet type, which handles both IPv4 and IPv6 addresses, we
should use that instead. Where this is done elsewhere in the codebase,
the column is called ip_address, so we should follow that convention as
well.
This migration uses a SQL command to populate the new field from the old
one, so as not to rely on the TopicLinkClick model class, which should
keep the migration from failing if that class is modified in the future.
