From 1b2e2c76d66be9a01478f831051ab670982d70ff Mon Sep 17 00:00:00 2001
From: Abimael Martell <1450169+abimaelmartell@users.noreply.github.com>
Date: Wed, 24 Jun 2026 01:27:27 -0600
Subject: [PATCH] fix(extractor): make trace previews unicode-safe (#113)

---
 src/extractor/layout.rs |  6 +-----
 src/extractor/mod.rs    | 23 ++++++++++++++++++-----
 2 files changed, 19 insertions(+), 10 deletions(-)

diff --git a/src/extractor/layout.rs b/src/extractor/layout.rs
index 06ae3c5..0a4530e 100644
--- a/src/extractor/layout.rs
+++ b/src/extractor/layout.rs
@@ -1234,11 +1234,7 @@ pub(crate) fn group_into_lines_with_thresholds(
                             ci,
                             item.x,
                             item.y,
-                            if item.text.len() > 60 {
-                                &item.text[..60]
-                            } else {
-                                &item.text
-                            }
+                            super::trace_text_preview(&item.text, 60)
                         );
                     }
                 }
diff --git a/src/extractor/mod.rs b/src/extractor/mod.rs
index a657332..2f293d9 100644
--- a/src/extractor/mod.rs
+++ b/src/extractor/mod.rs
@@ -33,6 +33,13 @@ pub(crate) use layout::ColumnRegion;
 // Public API
 // ---------------------------------------------------------------------------
 
+pub(crate) fn trace_text_preview(text: &str, max_chars: usize) -> &str {
+    match text.char_indices().nth(max_chars) {
+        Some((idx, _)) => &text[..idx],
+        None => text,
+    }
+}
+
 /// Extract text from PDF file as plain string
 pub fn extract_text<P: AsRef<Path>>(path: P) -> Result<String, PdfError> {
     crate::validate_pdf_file(&path)?;
@@ -195,11 +202,7 @@ fn extract_positioned_text_impl(
                     item.width,
                     item.font_size,
                     item.font,
-                    if item.text.len() > 80 {
-                        &item.text[..80]
-                    } else {
-                        &item.text
-                    }
+                    trace_text_preview(&item.text, 80)
                 );
             }
         }
@@ -574,6 +577,16 @@ mod tests {
         }
     }
 
+    #[test]
+    fn trace_text_preview_truncates_on_char_boundary() {
+        let text = format!("{}{}tail", "a".repeat(79), '\u{FFFD}');
+        let preview = trace_text_preview(&text, 80);
+
+        assert_eq!(preview.chars().count(), 80);
+        assert!(text.is_char_boundary(preview.len()));
+        assert!(preview.ends_with('\u{FFFD}'));
+    }
+
     #[test]
     fn merge_items_no_space_before_period() {
         // Simulate Tc/Tw-adjusted width: "date" width is smaller than the gap