From 2c28a26adcc381037f24a936b34e0e6b1f2109ac Mon Sep 17 00:00:00 2001
From: David Raznick <kindly@gmail.com>
Date: Thu, 25 Jan 2024 21:48:11 +0000
Subject: [PATCH] fix truncate in middle of unicode char

---
 src/lib.rs | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/src/lib.rs b/src/lib.rs
index 4f90e30..1c6d33f 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -2587,7 +2587,11 @@ impl FlatFiles {
 
                     if cell.len() > 32767 {
                         log::warn!("WARNING: Cell larger than 32767 chararcters which is too large for XLSX format. The cell will be truncated, so some data will be missing.");
-                        cell.truncate(32767)
+                        let mut index: usize = 32767;
+                        while !cell.is_char_boundary(index) {
+                            index -= 1;
+                        }
+                        cell.truncate(index)
                     }
 
                     if metadata.describers[order].guess_type().0 == "number" {
@@ -4376,7 +4380,7 @@ mod tests {
             json!({}),
         )
     }
-
+ 
     #[test]
     fn test_s3_input() {
         if std::env::var("AWS_DEFAULT_REGION").is_ok() {