bendns' repos / helix
Unnamed repository; edit this file 'description' to name the repository.
Diffstat (limited to 'helix-stdx/src/rope.rs')
-rw-r--r--helix-stdx/src/rope.rs62
1 files changed, 62 insertions, 0 deletions
diff --git a/helix-stdx/src/rope.rs b/helix-stdx/src/rope.rs
index 7e2549f5..2695555e 100644
--- a/helix-stdx/src/rope.rs
+++ b/helix-stdx/src/rope.rs
@@ -3,6 +3,7 @@ use std::ops::{Bound, RangeBounds};
pub use regex_cursor::engines::meta::{Builder as RegexBuilder, Regex};
pub use regex_cursor::regex_automata::util::syntax::Config;
use regex_cursor::{Input as RegexInput, RopeyCursor};
+use ropey::str_utils::byte_to_char_idx;
use ropey::RopeSlice;
pub trait RopeSliceExt<'a>: Sized {
@@ -16,6 +17,23 @@ pub trait RopeSliceExt<'a>: Sized {
fn regex_input_at<R: RangeBounds<usize>>(self, char_range: R) -> RegexInput<RopeyCursor<'a>>;
fn first_non_whitespace_char(self) -> Option<usize>;
fn last_non_whitespace_char(self) -> Option<usize>;
+ /// returns the char idx of `byte_idx`, if `byte_idx` is a char boundary
+ /// this function behaves the same as `byte_to_char` but if `byte_idx` is
+ /// not a valid char boundary (so within a char) this will return the next
+ /// char index.
+ ///
+ /// # Example
+ ///
+ /// ```
+ /// # use ropey::RopeSlice;
+ /// # use helix_stdx::rope::RopeSliceExt;
+ /// let text = RopeSlice::from("😆");
+ /// for i in 1..text.len_bytes() {
+ /// assert_eq!(text.byte_to_char(i), 0);
+ /// assert_eq!(text.byte_to_next_char(i), 1);
+ /// }
+ /// ```
+ fn byte_to_next_char(self, byte_idx: usize) -> usize;
}
impl<'a> RopeSliceExt<'a> for RopeSlice<'a> {
@@ -75,4 +93,48 @@ impl<'a> RopeSliceExt<'a> for RopeSlice<'a> {
.position(|ch| !ch.is_whitespace())
.map(|pos| self.len_chars() - pos - 1)
}
+
+ /// returns the char idx of `byte_idx`, if `byte_idx` is
+ /// a char boundary this function behaves the same as `byte_to_char`
+ fn byte_to_next_char(self, mut byte_idx: usize) -> usize {
+ let (chunk, chunk_byte_off, chunk_char_off, _) = self.chunk_at_byte(byte_idx);
+ byte_idx -= chunk_byte_off;
+ let is_char_boundary =
+ is_utf8_char_boundary(chunk.as_bytes().get(byte_idx).copied().unwrap_or(0));
+ chunk_char_off + byte_to_char_idx(chunk, byte_idx) + !is_char_boundary as usize
+ }
+}
+
+// copied from std
+#[inline]
+const fn is_utf8_char_boundary(b: u8) -> bool {
+ // This is bit magic equivalent to: b < 128 || b >= 192
+ (b as i8) >= -0x40
+}
+
+#[cfg(test)]
+mod tests {
+ use ropey::RopeSlice;
+
+ use crate::rope::RopeSliceExt;
+
+ #[test]
+ fn next_char_at_byte() {
+ for i in 0..=6 {
+ assert_eq!(RopeSlice::from("foobar").byte_to_next_char(i), i);
+ }
+ for char_idx in 0..10 {
+ let len = "😆".len();
+ assert_eq!(
+ RopeSlice::from("😆😆😆😆😆😆😆😆😆😆").byte_to_next_char(char_idx * len),
+ char_idx
+ );
+ for i in 1..=len {
+ assert_eq!(
+ RopeSlice::from("😆😆😆😆😆😆😆😆😆😆").byte_to_next_char(char_idx * len + i),
+ char_idx + 1
+ );
+ }
+ }
+ }
}