michael@0: /*
michael@0:  **********************************************************************
michael@0:  *   Copyright (C) 2005-2008, International Business Machines
michael@0:  *   Corporation and others.  All Rights Reserved.
michael@0:  **********************************************************************
michael@0:  */
michael@0: 
michael@0: #ifndef __INPUTEXT_H
michael@0: #define __INPUTEXT_H
michael@0: 
michael@0: /**
michael@0:  * \file
michael@0:  * \internal
michael@0:  *
michael@0:  * This is an internal header for the Character Set Detection code. The
michael@0:  * name is probably too generic...
michael@0:  */
michael@0: 
michael@0: 
michael@0: #include "unicode/uobject.h"
michael@0: 
michael@0: #if !UCONFIG_NO_CONVERSION
michael@0: 
michael@0: U_NAMESPACE_BEGIN 
michael@0: 
michael@0: class InputText : public UMemory
michael@0: {
michael@0:     // Prevent copying
michael@0:     InputText(const InputText &);
michael@0: public:
michael@0:     InputText(UErrorCode &status);
michael@0:     ~InputText();
michael@0: 
michael@0:     void setText(const char *in, int32_t len);
michael@0:     void setDeclaredEncoding(const char *encoding, int32_t len);
michael@0:     UBool isSet() const; 
michael@0:     void MungeInput(UBool fStripTags);
michael@0: 
michael@0:     // The text to be checked.  Markup will have been
michael@0:     //   removed if appropriate.
michael@0:     uint8_t    *fInputBytes;
michael@0:     int32_t     fInputLen;          // Length of the byte data in fInputBytes.
michael@0:     // byte frequency statistics for the input text.
michael@0:     //   Value is percent, not absolute.
michael@0:     //   Value is rounded up, so zero really means zero occurences. 
michael@0:     int16_t  *fByteStats;
michael@0:     UBool     fC1Bytes;          // True if any bytes in the range 0x80 - 0x9F are in the input;false by default
michael@0:     char     *fDeclaredEncoding;
michael@0: 
michael@0:     const uint8_t           *fRawInput;     // Original, untouched input bytes.
michael@0:     //  If user gave us a byte array, this is it.
michael@0:     //  If user gave us a stream, it's read to a 
michael@0:     //   buffer here.
michael@0:     int32_t                  fRawLength;    // Length of data in fRawInput array.
michael@0: 
michael@0: };
michael@0: 
michael@0: U_NAMESPACE_END
michael@0: 
michael@0: #endif
michael@0: #endif /* __INPUTEXT_H */