diff options
Diffstat (limited to 'src/util')
| -rw-r--r-- | src/util/normalize_identifier.rs | 35 |
1 files changed, 33 insertions, 2 deletions
diff --git a/src/util/normalize_identifier.rs b/src/util/normalize_identifier.rs index 870fd33..c287e1a 100644 --- a/src/util/normalize_identifier.rs +++ b/src/util/normalize_identifier.rs @@ -1,6 +1,37 @@ -//! To do. +//! Utility to normalize identifiers. -/// To do. +/// Normalize an identifier, as found in references and +/// [definitions][definition], so it can be compared when matching. +/// +/// This collapsed whitespace found in markdown (`\t`, `\r`, `\n`, and ` `) +/// into one space, trims it (as in, dropping the first and last space), +/// and then performs unicode case folding twice: first by uppercasing +/// lowercase characters, and then lowercasing uppercase characters. +/// +/// Some characters are considered “uppercase”, such as U+03F4 (`ϴ`), but if +/// their lowercase counterpart (U+03B8 (`θ`)) is uppercased will result in a +/// different uppercase character (U+0398 (`Θ`)). +/// Hence, to get that form, we perform both upper- and lowercase. +/// +/// ## Examples +/// +/// ```rust ignore +/// micromark::util::normalize_identifier::normalize_identifier; +/// +/// assert_eq!(normalize_identifier(" a "), "a"); +/// assert_eq!(normalize_identifier("a\t\r\nb"), "a b"); +/// assert_eq!(normalize_identifier("ПРИВЕТ"), "привет"); +/// assert_eq!(normalize_identifier("Привет"), "привет"); +/// assert_eq!(normalize_identifier("привет"), "привет"); +/// ``` +/// +/// ## References +/// +/// * [`micromark-util-normalize-identifier` in `micromark`](https://github.com/micromark/micromark/tree/main/packages/micromark-util-normalize-identifier) +/// +/// [definition]: crate::construct::definition +/// +/// <!--To do: link resource.--> pub fn normalize_identifier(value: &str) -> String { let mut codes = vec![]; let mut at_start = true; |