Thoughts on clang-format migration

[scarf005]

Reasoning

• astyle is outdated
• does not support c++20 language features (spaceship operator <=> is split into <= >)
• does not work consistently on new C++ features (such as: control statement inside lambda)

Pros

• Easier to integrate (default in many C++ extensions)
• Consistent behavior

Cons

• Some styles may be impossible to port over
• Large portion of code may be needed to be formatted
• Needs to change every workflow
• Porting from DDA would be a nightmare

Conclusion

• The cost for migration outweighs benefits.
• Halted until at least 99% of features can be ported over

Example

BasedOnStyle: Google

# One True Brace Style
BreakBeforeBraces: Custom
BraceWrapping:
    AfterClass: true
    AfterControlStatement: Never
    AfterNamespace: true
    AfterStruct: true
    AfterUnion: true
    AfterFunction: true
    AfterExternBlock: true
    BeforeCatch: true
    BeforeElse: true
    IndentBraces: false

IndentWidth: 4
SortIncludes: Never

# maximum length for a single unbroken line
ColumnLimit: 100

SpaceAfterTemplateKeyword: false

AllowShortFunctionsOnASingleLine: None

IndentAccessModifiers: true

UseTab: Never

SpacesInParentheses: true
SpacesInSquareBrackets: true
SpaceBeforeParens: Never
SpacesBeforeTrailingComments: 1

[scarf005]

it's discussed in other PR (which i don't remember), but using clang-format would make porting from DDA impossible. therefore, we should use astyle as-is.