ICU-59180.0.1.tar.gz

[apple/icu.git] / icuSources / i18n / csrutf8.cpp

diff --git a/icuSources/i18n/csrutf8.cpp b/icuSources/i18n/csrutf8.cpp
index 8af5307bdc39faf458e00b779767c23e51e6d47a..b42bd8b39e65cbc87c208b2353ed729536edd3af 100644 (file)
--- a/icuSources/i18n/csrutf8.cpp
+++ b/icuSources/i18n/csrutf8.cpp
@@ -1,6 +1,8 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
 /*
  **********************************************************************
- *   Copyright (C) 2005-2006, International Business Machines
+ *   Copyright (C) 2005-2014, International Business Machines
  *   Corporation and others.  All Rights Reserved.
  **********************************************************************
  */
@@ -10,6 +12,7 @@
 #if !UCONFIG_NO_CONVERSION
 
 #include "csrutf8.h"
+#include "csmatch.h"
 
 U_NAMESPACE_BEGIN
 
@@ -23,88 +26,85 @@ const char *CharsetRecog_UTF8::getName() const
     return "UTF-8";
 }
 
-int32_t CharsetRecog_UTF8::match(InputText* det) {
+UBool CharsetRecog_UTF8::match(InputText* input, CharsetMatch *results) const {
     bool hasBOM = FALSE;
     int32_t numValid = 0;
     int32_t numInvalid = 0;
-    const uint8_t *input = det->fRawInput;
+    const uint8_t *inputBytes = input->fRawInput;
     int32_t i;
     int32_t trailBytes = 0;
     int32_t confidence;
 
-    if (det->fRawLength >= 3 && 
-        input[0] == 0xEF && input[1] == 0xBB && input[2] == 0xBF) {
+    if (input->fRawLength >= 3 && 
+        inputBytes[0] == 0xEF && inputBytes[1] == 0xBB && inputBytes[2] == 0xBF) {
             hasBOM = TRUE;
-        }
+    }
 
-        // Scan for multi-byte sequences
-        for (i=0; i < det->fRawLength; i += 1) {
-            int32_t b = input[i];
+    // Scan for multi-byte sequences
+    for (i=0; i < input->fRawLength; i += 1) {
+        int32_t b = inputBytes[i];
 
-            if ((b & 0x80) == 0) {
-                continue;   // ASCII
-            }
+        if ((b & 0x80) == 0) {
+            continue;   // ASCII
+        }
 
-            // Hi bit on char found.  Figure out how long the sequence should be
-            if ((b & 0x0E0) == 0x0C0) {
-                trailBytes = 1;                
-            } else if ((b & 0x0F0) == 0x0E0) {
-                trailBytes = 2;
-            } else if ((b & 0x0F8) == 0xF0) {
-                trailBytes = 3;
-            } else {
-                numInvalid += 1;
+        // Hi bit on char found.  Figure out how long the sequence should be
+        if ((b & 0x0E0) == 0x0C0) {
+            trailBytes = 1;
+        } else if ((b & 0x0F0) == 0x0E0) {
+            trailBytes = 2;
+        } else if ((b & 0x0F8) == 0xF0) {
+            trailBytes = 3;
+        } else {
+            numInvalid += 1;
+            continue;
+        }
 
-                if (numInvalid > 5) {
-                    break;
-                }
+        // Verify that we've got the right number of trail bytes in the sequence
+        for (;;) {
+            i += 1;
 
-                trailBytes = 0;
+            if (i >= input->fRawLength) {
+                break;
             }
 
-            // Verify that we've got the right number of trail bytes in the sequence
-            for (;;) {
-                i += 1;
-
-                if (i >= det->fRawLength) {
-                    break;
-                }
+            b = inputBytes[i];
 
-                b = input[i];
-
-                if ((b & 0xC0) != 0x080) {
-                    numInvalid += 1;
-                    break;
-                }
-
-                if (--trailBytes == 0) {
-                    numValid += 1;
-                    break;
-                }
+            if ((b & 0xC0) != 0x080) {
+                numInvalid += 1;
+                break;
             }
 
+            if (--trailBytes == 0) {
+                numValid += 1;
+                break;
+            }
         }
 
-        // Cook up some sort of confidence score, based on presense of a BOM
-        //    and the existence of valid and/or invalid multi-byte sequences.
-        confidence = 0;
-        if (hasBOM && numInvalid == 0) {
-            confidence = 100;
-        } else if (hasBOM && numValid > numInvalid*10) {
-            confidence = 80;
-        } else if (numValid > 3 && numInvalid == 0) {
-            confidence = 100;            
-        } else if (numValid > 0 && numInvalid == 0) {
-            confidence = 80;
-        } else if (numValid == 0 && numInvalid == 0) {
-            // Plain ASCII.  
-            confidence = 10;            
-        } else if (numValid > numInvalid*10) {
-            // Probably corruput utf-8 data.  Valid sequences aren't likely by chance.
-            confidence = 25;
-        }
-
-        return confidence;
+    }
+
+    // Cook up some sort of confidence score, based on presence of a BOM
+    //    and the existence of valid and/or invalid multi-byte sequences.
+    confidence = 0;
+    if (hasBOM && numInvalid == 0) {
+        confidence = 100;
+    } else if (hasBOM && numValid > numInvalid*10) {
+        confidence = 80;
+    } else if (numValid > 3 && numInvalid == 0) {
+        confidence = 100;
+    } else if (numValid > 0 && numInvalid == 0) {
+        confidence = 80;
+    } else if (numValid == 0 && numInvalid == 0) {
+        // Plain ASCII. Confidence must be > 10, it's more likely than UTF-16, which
+        //              accepts ASCII with confidence = 10.
+        confidence = 15;
+    } else if (numValid > numInvalid*10) {
+        // Probably corruput utf-8 data.  Valid sequences aren't likely by chance.
+        confidence = 25;
+    }
+
+    results->set(input, this, confidence);
+    return (confidence > 0);
 }
 
 U_NAMESPACE_END