Fix false negative with HZ detection (#129)

* Fix false negative with HZ detection * Update
CharsetDetector · Oct 26, 2021 · 975bc42 · 975bc42
1 parent 1f53769
commit 975bc42
Show file tree

Hide file tree

Showing 2 changed files with 17 additions and 0 deletions.
diff --git a/src/CharsetDetector.cs b/src/CharsetDetector.cs
@@ -470,6 +470,10 @@ private DetectionResult DataEnd()
                 //TODO why done isn't true?
                 return new DetectionResult(new DetectionDetail(CodepageName.ASCII, 1.0f));
             }
+            else if (InputState == InputState.EscASCII)
+            {
+                return new DetectionResult(new DetectionDetail(CodepageName.ASCII, 1.0f));
+            }
 
             return new DetectionResult();
         }

diff --git a/tests/CharsetDetectorTest.cs b/tests/CharsetDetectorTest.cs
@@ -34,6 +34,19 @@ public void TestAscii()
             }
         }
 
+        [Test]
+        public void TestAscii_with_HZ_sequence()
+        {
+            const string text = "virtual ~{{NETCLASS_NAME}}();";
+            var stream = AsciiToStream(text);
+            using (stream)
+            {
+                var result = CharsetDetector.DetectFromStream(stream);
+                Assert.AreEqual(CodepageName.ASCII, result.Detected.EncodingName);
+                Assert.AreEqual(1.0f, result.Detected.Confidence);
+            }
+        }
+
         private static MemoryStream AsciiToStream(string s)
         {
             return new MemoryStream(Encoding.ASCII.GetBytes(s));