ui: word break with grapheme length, not bytes

​#69 East characters are not working.

4 files changed, 42 insertions, 31 deletions

diff --git a/ui/src/terminal/cells.rs b/ui/src/terminal/cells.rs
index 845900a1..efd93fb1 100644
--- a/ui/src/terminal/cells.rs
+++ b/ui/src/terminal/cells.rs
@@ -24,6 +24,7 @@
  colors and attributes.
 */
 use super::position::*;
+use super::grapheme_clusters::*;
 use std::convert::From;
 use std::fmt;
 use std::ops::{Deref, DerefMut, Index, IndexMut};
@@ -733,7 +734,6 @@ pub fn write_string_to_grid(
     (x, y)
 }
 
-// TODO UTF-8 incompatible
 pub fn word_break_string(mut s: &str, width: usize) -> Vec<&str> {
     let mut ret: Vec<&str> = Vec::with_capacity(16);
     loop {
@@ -753,12 +753,16 @@ pub fn word_break_string(mut s: &str, width: usize) -> Vec<&str> {
                 continue;
             }
         }
-        if s.len() > width {
-            if let Some(next_idx) = s.as_bytes()[..width]
+        let graphemes = s.graphemes_indices();
+        if graphemes.len() > width {
+            // use grapheme indices and find position of " " graphemes
+            if let Some(next_idx) = graphemes[..width]
                 .iter()
-                .rposition(u8::is_ascii_whitespace)
+                .rposition(|(_, g)| *g == " ")
             {
+                let next_idx = graphemes[next_idx].0;
                 ret.push(&s[..next_idx]);
+                eprintln!("width = {} w = {} l = {:?}\n\n", width, ret.last().unwrap().grapheme_width(), ret.last().unwrap());
                 s = &s[next_idx + 1..];
             } else {
                 ret.push(&s[..width]);
diff --git a/ui/src/terminal/grapheme_clusters.rs b/ui/src/terminal/grapheme_clusters.rs
index 4e661c9b..2135ce8f 100644
--- a/ui/src/terminal/grapheme_clusters.rs
+++ b/ui/src/terminal/grapheme_clusters.rs
@@ -8,34 +8,42 @@
 
 */
 
-use super::wcwidth::{wcwidth, CodePointsIter};
 use super::*;
 
-pub fn split_graphemes(s: &str) -> Vec<&str> {
-    UnicodeSegmentation::graphemes(s, true).collect::<Vec<&str>>()
-}
 
-pub fn next_grapheme(s: &str) -> Option<(usize, &str)> {
-    UnicodeSegmentation::grapheme_indices(s, true).next()
-}
+pub trait Graphemes: UnicodeSegmentation + CodePointsIter {
+    fn split_graphemes<'a>(&'a self) -> Vec<&'a str> {
+        UnicodeSegmentation::graphemes(self, true).collect::<Vec<&str>>()
+    }
 
-pub fn last_grapheme(s: &str) -> Option<(usize, &str)> {
-    UnicodeSegmentation::grapheme_indices(s, true).next_back()
-}
+    fn graphemes_indices<'a>(&'a self) -> Vec<(usize, &'a str)> {
+        UnicodeSegmentation::grapheme_indices(self, true).collect::<Vec<(usize, &str)>>()
+    }
 
-pub fn grapheme_width(grapheme: &str) -> i32 {
-    let mut count = 0;
-    for c in grapheme.code_points() {
-        count += if let Some(c) = wcwidth(c) {
-            c as i32
-        } else {
-            -1
-        };
+    fn next_grapheme<'a>(&'a self) -> Option<(usize, &'a str)> {
+        UnicodeSegmentation::grapheme_indices(self, true).next()
     }
 
-    count
+    fn last_grapheme<'a>(&'a self) -> Option<(usize, &'a str)> {
+        UnicodeSegmentation::grapheme_indices(self, true).next_back()
+    }
+
+    fn grapheme_width(&self) -> i32 {
+        let mut count = 0;
+        for c in self.code_points() {
+            count += if let Some(c) = wcwidth(c) {
+                c as i32
+            } else {
+                -1
+            };
+        }
+
+        count
+    }
 }
 
+impl Graphemes for str {}
+
 //#[derive(PartialEq)]
 //enum Property {
 //    CR,
diff --git a/ui/src/terminal/text_editing.rs b/ui/src/terminal/text_editing.rs
index 326ba1aa..6ce24fe2 100644
--- a/ui/src/terminal/text_editing.rs
+++ b/ui/src/terminal/text_editing.rs
@@ -22,7 +22,7 @@ impl UText {
         }
 
         let (first, _) = self.content.split_at(cursor_pos);
-        self.grapheme_cursor_pos = split_graphemes(first).len();
+        self.grapheme_cursor_pos = first.split_graphemes().len();
         self.cursor_pos = cursor_pos;
     }
 
@@ -34,7 +34,7 @@ impl UText {
         self.content
     }
     pub fn grapheme_len(&self) -> usize {
-        split_graphemes(&self.content).len()
+        self.content.split_graphemes().len()
     }
 
     pub fn cursor_inc(&mut self) {
@@ -42,8 +42,8 @@ impl UText {
             return;
         }
 
-        let (_, right) = std::dbg!(self.content.split_at(self.cursor_pos));
-        if let Some((_, graph)) = std::dbg!(next_grapheme(right)) {
+        let (_, right) = self.content.split_at(self.cursor_pos);
+        if let Some((_, graph)) = right.next_grapheme() {
             self.cursor_pos += graph.len();
             self.grapheme_cursor_pos += 1;
         }
@@ -52,8 +52,8 @@ impl UText {
         if self.cursor_pos == 0 {
             return;
         }
-        let (left, _) = std::dbg!(self.content.split_at(self.cursor_pos));
-        if let Some((_, graph)) = std::dbg!(last_grapheme(left)) {
+        let (left, _) = self.content.split_at(self.cursor_pos);
+        if let Some((_, graph)) = left.last_grapheme() {
             self.cursor_pos -= graph.len();
             self.grapheme_cursor_pos -= 1;
         }
@@ -116,7 +116,7 @@ impl UText {
              * left = xxxxxx....xxgg;
              * right = xxx;
              */
-            if let Some((offset, graph)) = std::dbg!(last_grapheme(left)) {
+            if let Some((offset, graph)) = left.last_grapheme() {
                 (offset, graph.len())
             } else {
                 return;
diff --git a/ui/src/terminal/wcwidth.rs b/ui/src/terminal/wcwidth.rs
index f0493b7d..b6202102 100644
--- a/ui/src/terminal/wcwidth.rs
+++ b/ui/src/terminal/wcwidth.rs
@@ -41,7 +41,6 @@ impl<'a> Iterator for CodePointsIterator<'a> {
     type Item = WChar;
 
     fn next(&mut self) -> Option<WChar> {
-        println!("rest = {:?}", self.rest);
         if self.rest.is_empty() {
             return None;
         }