fix truncate unicode grapheme clusters

3 files changed, 28 insertions, 5 deletions

diff --git a/Cargo.lock b/Cargo.lock
index 2fed4cc..b9a4a6c 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -13,6 +13,7 @@ dependencies = [
  "getopts 0.2.15 (registry+https://github.com/rust-lang/crates.io-index)",
  "regex 0.2.10 (registry+https://github.com/rust-lang/crates.io-index)",
  "terminal_size 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)",
+ "unicode-segmentation 1.2.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "unicode-width 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
@@ -93,6 +94,11 @@ version = "0.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 
 [[package]]
+name = "unicode-segmentation"
+version = "1.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+
+[[package]]
 name = "unicode-width"
 version = "0.1.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -137,6 +143,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 "checksum terminal_size 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)" = "ef4f7fdb2a063032d361d9a72539380900bc3e0cd9ffc9ca8b677f8c855bae0f"
 "checksum thread_local 0.3.5 (registry+https://github.com/rust-lang/crates.io-index)" = "279ef31c19ededf577bfd12dfae728040a21f635b06a24cd670ff510edd38963"
 "checksum ucd-util 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)" = "fd2be2d6639d0f8fe6cdda291ad456e23629558d466e2789d2c3e9892bda285d"
+"checksum unicode-segmentation 1.2.0 (registry+https://github.com/rust-lang/crates.io-index)" = "a8083c594e02b8ae1654ae26f0ade5158b119bd88ad0e8227a5d8fcd72407946"
 "checksum unicode-width 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)" = "bf3a113775714a22dcb774d8ea3655c53a32debae63a063acc00a91cc586245f"
 "checksum unreachable 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)" = "382810877fe448991dfc7f0dd6e3ae5d58088fd0ea5e35189655f84e6814fa56"
 "checksum utf8-ranges 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)" = "662fab6525a98beff2921d7f61a39e7d59e0b425ebc7d0d9e66d316e55124122"
diff --git a/Cargo.toml b/Cargo.toml
index 6dc6d4e..e7aecea 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -1,6 +1,6 @@
 [package]
 name = "dutree"
-version = "0.2.4"
+version = "0.2.5"
 authors = ["nacho <nacho@ownyourbits.com>"]
 description = "Command line tool to analyze disk usage"
 repository = "https://github.com/nachoparker/dutree"
@@ -14,3 +14,4 @@ getopts = "0.2"
 terminal_size = "0.1.7"
 regex = "0.2"
 unicode-width = "0.1.1"
+unicode-segmentation = "1.2.0"
diff --git a/src/lib.rs b/src/lib.rs
index 4915502..b1fcae7 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -35,6 +35,10 @@
 //!
 
 extern crate unicode_width;
+use unicode_width::UnicodeWidthStr;
+
+extern crate unicode_segmentation;
+use unicode_segmentation::UnicodeSegmentation;
 
 extern crate getopts;
 use getopts::Options;
@@ -324,10 +328,21 @@ impl<'a> Entry<'a> {
                 let tree_width = (open_parents.len() + 1) * 3; // 3 chars per tree branch
                 if tree_name_width >= tree_width {
                     let name_width  = tree_name_width - tree_width;
-                    let length = unicode_width::UnicodeWidthStr::width(entry.name.as_str());
-
-                    let mut name = entry.name.clone();
-                    name.truncate( name_width );
+                    let length = UnicodeWidthStr::width(entry.name.as_str());
+
+                    // truncate Unicode string to name_width
+                    let graphemes = UnicodeSegmentation::graphemes( entry.name.as_str(), true );
+                    let mut i   = 0;
+                    let mut vec = Vec::new();
+                    for cluster in graphemes {
+                        let w = UnicodeWidthStr::width( cluster );
+                        if i + w <= name_width {
+                            i += w;
+                            vec.push( cluster );
+                        }
+                    }
+                    let mut name = String::new();
+                    vec.iter().for_each( |cluster| name.push_str( cluster ) );
 
                     // surround name by ANSII color escape sequences
                     if let Some( ref col_str ) = entry.color {