diff --git a/judge/dispatcher.py b/judge/dispatcher.py
index c4480956c..75e63e4cb 100644
--- a/judge/dispatcher.py
+++ b/judge/dispatcher.py
@@ -164,6 +164,10 @@ def judge(self):
             resp["data"].sort(key=lambda x: int(x["test_case"]))
             self.submission.info = resp
             self._compute_statistic_info(resp["data"])
+            if self.problem.pe_ignored:
+                for case in resp["data"]:
+                    if case["result"] is JudgeStatus.PRESNTATION_ERROR:
+                        case["result"] = 0
             error_test_case = list(filter(lambda case: case["result"] != 0, resp["data"]))
             # ACM模式下,多个测试点全部正确则AC，否则取第一个错误的测试点的状态
             # OI模式下, 若多个测试点全部正确则AC， 若全部错误则取第一个错误测试点状态，否则为部分正确
diff --git a/problem/migrations/0013_auto_20181227_0719.py b/problem/migrations/0013_auto_20181227_0719.py
new file mode 100644
index 000000000..9be46b53d
--- /dev/null
+++ b/problem/migrations/0013_auto_20181227_0719.py
@@ -0,0 +1,45 @@
+# -*- coding: utf-8 -*-
+# Generated by Django 1.11.4 on 2018-12-27 07:19
+from __future__ import unicode_literals
+
+from django.db import migrations, models
+from django.conf import settings
+import os
+import json
+import hashlib
+
+
+def update_all_stripped_md5(apps, schema_editor):
+    Problem = apps.get_model("problem", "Problem")
+    problems = Problem.objects.all()
+    for problem in problems:
+        test_case_dir = os.path.join(settings.TEST_CASE_DIR, problem.test_case_id)
+        try:
+            with open(os.path.join(test_case_dir, "info")) as f:
+                info = json.load(f)
+            if not info["spj"]:
+                for id in info["test_cases"]:
+                    with open(os.path.join(test_case_dir,f"{id}.out"),"rb") as f:
+                        content = f.read()
+                        stripped_md5 = hashlib.md5(b"\n".join([x.rstrip() for x in content.split(b"\n") if len(x) > 0])).hexdigest()
+                        info["test_cases"][id]["all_stripped_output_md5"] = stripped_md5
+                    with open(os.path.join(test_case_dir, "info"), "w", encoding="utf-8") as f:
+                        f.write(json.dumps(info, indent=4))
+        except Exception as err:
+            print(f"{test_case_dir} info not found")
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('problem', '0012_auto_20180501_0436'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='problem',
+            name='pe_ignored',
+            field=models.BooleanField(default=False),
+        ),
+        migrations.RunPython(update_all_stripped_md5, reverse_code=migrations.RunPython.noop),
+    ]
diff --git a/problem/models.py b/problem/models.py
index 29e317bd3..61e525c99 100644
--- a/problem/models.py
+++ b/problem/models.py
@@ -45,6 +45,7 @@ class Problem(models.Model):
     languages = JSONField()
     template = JSONField()
     create_time = models.DateTimeField(auto_now_add=True)
+    pe_ignored = models.BooleanField(default=False)
     # we can not use auto_now here
     last_update_time = models.DateTimeField(null=True)
     created_by = models.ForeignKey(User)
diff --git a/problem/serializers.py b/problem/serializers.py
index dbea96b7d..7e05f6fad 100644
--- a/problem/serializers.py
+++ b/problem/serializers.py
@@ -48,6 +48,7 @@ class CreateOrEditProblemSerializer(serializers.Serializer):
     spj_code = serializers.CharField(allow_blank=True, allow_null=True)
     spj_compile_ok = serializers.BooleanField(default=False)
     visible = serializers.BooleanField()
+    pe_ignored = serializers.BooleanField()
     difficulty = serializers.ChoiceField(choices=Difficulty.choices())
     tags = serializers.ListField(child=serializers.CharField(max_length=32), allow_empty=False)
     hint = serializers.CharField(allow_blank=True, allow_null=True)
diff --git a/problem/views/admin.py b/problem/views/admin.py
index a357cea5c..afa348f82 100644
--- a/problem/views/admin.py
+++ b/problem/views/admin.py
@@ -49,6 +49,7 @@ def process_zip(self, uploaded_zip_file, spj, dir=""):
 
         size_cache = {}
         md5_cache = {}
+        stripped_md5_cache = {}
 
         for item in test_case_list:
             with open(os.path.join(test_case_dir, item), "wb") as f:
@@ -56,6 +57,7 @@ def process_zip(self, uploaded_zip_file, spj, dir=""):
                 size_cache[item] = len(content)
                 if item.endswith(".out"):
                     md5_cache[item] = hashlib.md5(content.rstrip()).hexdigest()
+                    stripped_md5_cache[item] = hashlib.md5(b"\n".join([x.rstrip() for x in content.split(b"\n") if len(x) > 0])).hexdigest()
                 f.write(content)
         test_case_info = {"spj": spj, "test_cases": {}}
 
@@ -71,6 +73,7 @@ def process_zip(self, uploaded_zip_file, spj, dir=""):
             test_case_list = zip(*[test_case_list[i::2] for i in range(2)])
             for index, item in enumerate(test_case_list):
                 data = {"stripped_output_md5": md5_cache[item[1]],
+                        "all_stripped_output_md5": stripped_md5_cache[item[1]],
                         "input_size": size_cache[item[0]],
                         "output_size": size_cache[item[1]],
                         "input_name": item[0],
diff --git a/submission/models.py b/submission/models.py
index 261052e9d..d79f451ab 100644
--- a/submission/models.py
+++ b/submission/models.py
@@ -18,6 +18,7 @@ class JudgeStatus:
     PENDING = 6
     JUDGING = 7
     PARTIALLY_ACCEPTED = 8
+    PRESNTATION_ERROR = 9
 
 
 class Submission(models.Model):