From 5841856288e1e3278b3533ecd34da0c1e38d6472 Mon Sep 17 00:00:00 2001
From: Sathvik Gilakamsetty <spacetime1007@gmail.com>
Date: Sat, 16 May 2026 02:36:19 +0000
Subject: [PATCH] fix(streaming): fix SSE buffer corruption and add CJK
 sentence splitting (#1794)

---
 src/openhuman/channels/providers/presentation.rs     | 12 ++++++++++++
 .../channels/providers/presentation_tests.rs         |  9 +++++++++
 src/openhuman/providers/compatible_stream.rs         |  3 +--
 3 files changed, 22 insertions(+), 2 deletions(-)

diff --git a/src/openhuman/channels/providers/presentation.rs b/src/openhuman/channels/providers/presentation.rs
index bd2cd7136..e8c9e7758 100644
--- a/src/openhuman/channels/providers/presentation.rs
+++ b/src/openhuman/channels/providers/presentation.rs
@@ -311,6 +311,7 @@ fn split_sentences(text: &str) -> Vec<String> {
         current.push(chars[i]);
         let ch = chars[i];
 
+        // Latin sentence terminators: split on ". " followed by uppercase.
         if (ch == '.' || ch == '!' || ch == '?')
             && i + 2 < chars.len()
             && chars[i + 1] == ' '
@@ -325,6 +326,17 @@ fn split_sentences(text: &str) -> Vec<String> {
             continue;
         }
 
+        // CJK sentence terminators: split after fullwidth period/exclamation/question.
+        if (ch == '\u{3002}' || ch == '\u{FF01}' || ch == '\u{FF1F}') && i + 1 < chars.len() {
+            let trimmed = current.trim().to_string();
+            if !trimmed.is_empty() {
+                parts.push(trimmed);
+            }
+            current.clear();
+            i += 1;
+            continue;
+        }
+
         i += 1;
     }
 
diff --git a/src/openhuman/channels/providers/presentation_tests.rs b/src/openhuman/channels/providers/presentation_tests.rs
index e3e1dc70b..cb04ff0ac 100644
--- a/src/openhuman/channels/providers/presentation_tests.rs
+++ b/src/openhuman/channels/providers/presentation_tests.rs
@@ -180,6 +180,15 @@ fn split_sentences_single_sentence_without_terminator() {
     assert_eq!(out.len(), 1);
 }
 
+#[test]
+fn split_sentences_splits_on_cjk_terminators() {
+    let out = split_sentences("你好世界。今天天气很好！你觉得呢？");
+    assert_eq!(out.len(), 3);
+    assert_eq!(out[0], "你好世界。");
+    assert_eq!(out[1], "今天天气很好！");
+    assert_eq!(out[2], "你觉得呢？");
+}
+
 #[test]
 fn group_sentences_single_entry_roundtrip() {
     let v: Vec<String> = vec!["Hello world".into()];
diff --git a/src/openhuman/providers/compatible_stream.rs b/src/openhuman/providers/compatible_stream.rs
index 172ebe9ce..f19c53a0b 100644
--- a/src/openhuman/providers/compatible_stream.rs
+++ b/src/openhuman/providers/compatible_stream.rs
@@ -52,8 +52,7 @@ pub(crate) fn sse_bytes_to_chunks(
 
                     // Process complete lines
                     while let Some(pos) = buffer.find('\n') {
-                        let line = buffer.drain(..=pos).collect::<String>();
-                        buffer = buffer[pos + 1..].to_string();
+                        let line: String = buffer.drain(..=pos).collect();
 
                         match parse_sse_line(&line) {
                             Ok(Some(content)) => {