1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/intl/icu/source/i18n/csr2022.h Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,91 @@
1.4 +/*
1.5 + **********************************************************************
1.6 + * Copyright (C) 2005-2012, International Business Machines
1.7 + * Corporation and others. All Rights Reserved.
1.8 + **********************************************************************
1.9 + */
1.10 +
1.11 +#ifndef __CSR2022_H
1.12 +#define __CSR2022_H
1.13 +
1.14 +#include "unicode/utypes.h"
1.15 +
1.16 +#if !UCONFIG_NO_CONVERSION
1.17 +
1.18 +#include "csrecog.h"
1.19 +
1.20 +U_NAMESPACE_BEGIN
1.21 +
1.22 +class CharsetMatch;
1.23 +
1.24 +/**
1.25 + * class CharsetRecog_2022 part of the ICU charset detection imlementation.
1.26 + * This is a superclass for the individual detectors for
1.27 + * each of the detectable members of the ISO 2022 family
1.28 + * of encodings.
1.29 + *
1.30 + * The separate classes are nested within this class.
1.31 + *
1.32 + * @internal
1.33 + */
1.34 +class CharsetRecog_2022 : public CharsetRecognizer
1.35 +{
1.36 +
1.37 +public:
1.38 + virtual ~CharsetRecog_2022() = 0;
1.39 +
1.40 +protected:
1.41 +
1.42 + /**
1.43 + * Matching function shared among the 2022 detectors JP, CN and KR
1.44 + * Counts up the number of legal an unrecognized escape sequences in
1.45 + * the sample of text, and computes a score based on the total number &
1.46 + * the proportion that fit the encoding.
1.47 + *
1.48 + *
1.49 + * @param text the byte buffer containing text to analyse
1.50 + * @param textLen the size of the text in the byte.
1.51 + * @param escapeSequences the byte escape sequences to test for.
1.52 + * @return match quality, in the range of 0-100.
1.53 + */
1.54 + int32_t match_2022(const uint8_t *text,
1.55 + int32_t textLen,
1.56 + const uint8_t escapeSequences[][5],
1.57 + int32_t escapeSequences_length) const;
1.58 +
1.59 +};
1.60 +
1.61 +class CharsetRecog_2022JP :public CharsetRecog_2022
1.62 +{
1.63 +public:
1.64 + virtual ~CharsetRecog_2022JP();
1.65 +
1.66 + const char *getName() const;
1.67 +
1.68 + UBool match(InputText *textIn, CharsetMatch *results) const;
1.69 +};
1.70 +
1.71 +class CharsetRecog_2022KR :public CharsetRecog_2022 {
1.72 +public:
1.73 + virtual ~CharsetRecog_2022KR();
1.74 +
1.75 + const char *getName() const;
1.76 +
1.77 + UBool match(InputText *textIn, CharsetMatch *results) const;
1.78 +
1.79 +};
1.80 +
1.81 +class CharsetRecog_2022CN :public CharsetRecog_2022
1.82 +{
1.83 +public:
1.84 + virtual ~CharsetRecog_2022CN();
1.85 +
1.86 + const char* getName() const;
1.87 +
1.88 + UBool match(InputText *textIn, CharsetMatch *results) const;
1.89 +};
1.90 +
1.91 +U_NAMESPACE_END
1.92 +
1.93 +#endif
1.94 +#endif /* __CSR2022_H */
