From dd725466e5c28ad3e2d0872558571ae77e7931b0 Mon Sep 17 00:00:00 2001
From: Phuoc Dinh Le <47278241+cuom1999@users.noreply.github.com>
Date: Wed, 31 Mar 2021 21:46:42 -0500
Subject: [PATCH] Another attempt to fix VNese bugs with utf-8.

---
 judge/views/submission.py | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/judge/views/submission.py b/judge/views/submission.py
index 4e278ac..6c7d9e6 100644
--- a/judge/views/submission.py
+++ b/judge/views/submission.py
@@ -139,7 +139,16 @@ def group_test_cases(cases):
 
 def read_head_archive(archive, file):
     with archive.open(file) as f:
-        return f.read(settings.TESTCASE_VISIBLE_LENGTH + 3)
+        s = f.read(settings.TESTCASE_VISIBLE_LENGTH + 3)
+        # add this so there are no characters left behind (ex, 'á' = 2 utf-8 chars)
+        while True:
+            try:
+                s.decode('utf-8')
+                break
+            except UnicodeDecodeError:
+                s += f.read(1)
+        return s
+
 
 def get_visible_content(data):
     data = data or b''