Function gix_config::value::normalize

pub fn normalize(input: Cow<'_, BStr>) -> Cow<'_, BStr>

Removes quotes, if any, from the provided inputs, and transforms the 3 escape sequences \n, \t and \b into newline and tab respectively, while \b will remove the previous character.

It assumes the input contains a even number of unescaped quotes, and will unescape escaped quotes and everything else (even though the latter would have been rejected in the parsing stage).

The return values should be safe for value interpretation.

This has optimizations for fully-quoted values, where the returned value will be a borrowed reference if the only mutation necessary is to unquote the value.

This is the function used to normalize raw values from higher level abstractions. Generally speaking these high level abstractions will handle normalization for you, and you do not need to call this yourself. However, if you’re directly handling events from the parser, you may want to use this to help with value interpretation.

Generally speaking, you’ll want to use one of the variants of this function, such as normalize_bstr or normalize_bstring.

Examples

Values don’t need modification are returned borrowed, without allocation.

assert!(matches!(normalize_bstr("hello world"), Cow::Borrowed(_)));

Internally quoted values are turned into owned variant with quotes removed.

assert_eq!(normalize_bstr("hello \"world\""), Cow::<BStr>::Owned(BString::from("hello world")));

Escaped quotes are unescaped.

assert_eq!(normalize_bstr(r#"hello "world\"""#), Cow::<BStr>::Owned(BString::from(r#"hello world""#)));