…#1660)

* C# lexer: add keywords from newer language versions

 * `and` (pattern matching in C# 9)
 * `init` (records in C# 9)
 * `unmanaged` (new type constraint in C# 7)
 * `nint` (native sized integers in C# 9)
 * `nuint` (native sized integers in C# 9)
 * `not` (pattern matching in C# 9)
 * `notnull` (nullable reference types in C# 8)
 * `#nullable` (nullable reference types in C# 8)
 * `or` (pattern matching in C# 9)
 * `record` (records in C# 9)
 * `with` (records in C# 9)
 * `when` (exception filters in C# 6, pattern matching in C# 7)

* C# lexer: numeric literal improvements

 * add C# 7 binary literals, e.g. `0b0010`
 * add C# 7 digit separators in all numeric literals, e.g. `123_567_890`
 * add missing `M` type suffix designating `decimal`
 * make the exponent sign optional

…ws (#1911)

* Add Ruby 3.2 to CI build

* Bump power_assert to be compatible with Ruby 3.2

The new version removes TracePoint#binding check that causes
installation to fail in Ruby 3.2.

* Escape character that causes parsing issue

This ensures we escape `#` character to avoid parsing issue in Ruby 3.2.
This change is backward compatible with earlier versions.

* Update GitHub theme, add dark mode

This updates the GitHub theme and also adds support for a dark mode.

The color values are sourced from the primer/primitives repository.
https://github.com/primer/primitives

* Add text background color to GH theme

* Improve Python lexer

This includes three changes to the Python lexer:

1) Move the `decorator` rule before the `operator` rule.

   Before this change, decorators were never getting recognized. This
   was because if there was ever an `@`, it was always satisfying the
   `operator` rule. This remedies that by first checking for the
   `decorator` pattern and then the operator pattern.

2) Recognize functions and classes when they are called.

   Previously, functions and classes were only recognized when they were
   defined. With this change, they will also be recognized and styled
   when they are called.

3) Don't recognize imported modules as `Namespace`s.

   It is not desirable to highlight imported modules like namespaces.
   With this change, they are simply recognized as the general `Name`
   token.

* Add more python visual samples

* Add python class visual sample

---------

Co-authored-by: Tan Le <tan.le@hey.com>

* Porting Cisco IOS configuration lexer from pygments-routerlexers

* Fixing Cisco IOS sample for linter

* Recognize javascript functions

With this change, javascript functions will be recognized both when they
are defined and when they are called.

* Fix failing test

With the update to the javascript lexer, functions are now recognized.
This updates the test case to reflect this.

* Add function name token to regex to avoid adding state

* Support Unicode character function names

* Correctly recognize an anonymous function

* Recognize javascript classes

In order to recognize classes we have to tweak the `id` regex because we
want to try to differentiate between functions and classes. To recognize
classes, the first letter that appears in the name must not be a
lowercase character.

* Added javascript visual samples

Added for the new function and class (with unicode support) recognition.

* Add support for javascript class inheritance

This fixes a bug where closing bracket does not pops out of the function
state and causes subsequent functions not to be highlighted correctly.
This change ensures closing bracket in statements mixin is detecting
function state.