ICU 70.1  70.1
ucsdet.h
Go to the documentation of this file.
1 // © 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /*
4  **********************************************************************
5  * Copyright (C) 2005-2013, International Business Machines
6  * Corporation and others. All Rights Reserved.
7  **********************************************************************
8  * file name: ucsdet.h
9  * encoding: UTF-8
10  * indentation:4
11  *
12  * created on: 2005Aug04
13  * created by: Andy Heninger
14  *
15  * ICU Character Set Detection, API for C
16  *
17  * Draft version 18 Oct 2005
18  *
19  */
20 
21 #ifndef __UCSDET_H
22 #define __UCSDET_H
23 
24 #include "unicode/utypes.h"
25 
26 #if !UCONFIG_NO_CONVERSION
27 
28 #include "unicode/uenum.h"
29 
30 #if U_SHOW_CPLUSPLUS_API
31 #include "unicode/localpointer.h"
32 #endif // U_SHOW_CPLUSPLUS_API
33 
58 struct UCharsetDetector;
63 typedef struct UCharsetDetector UCharsetDetector;
64 
65 struct UCharsetMatch;
71 typedef struct UCharsetMatch UCharsetMatch;
72 
81 U_CAPI UCharsetDetector * U_EXPORT2
83 
93 U_CAPI void U_EXPORT2
95 
96 #if U_SHOW_CPLUSPLUS_API
97 
98 U_NAMESPACE_BEGIN
99 
110 
111 U_NAMESPACE_END
112 
113 #endif
114 
130 U_CAPI void U_EXPORT2
131 ucsdet_setText(UCharsetDetector *ucsd, const char *textIn, int32_t len, UErrorCode *status);
132 
133 
152 U_CAPI void U_EXPORT2
153 ucsdet_setDeclaredEncoding(UCharsetDetector *ucsd, const char *encoding, int32_t length, UErrorCode *status);
154 
155 
181 U_CAPI const UCharsetMatch * U_EXPORT2
183 
184 
215 U_CAPI const UCharsetMatch ** U_EXPORT2
216 ucsdet_detectAll(UCharsetDetector *ucsd, int32_t *matchesFound, UErrorCode *status);
217 
218 
219 
235 U_CAPI const char * U_EXPORT2
237 
261 U_CAPI int32_t U_EXPORT2
263 
293 U_CAPI const char * U_EXPORT2
295 
296 
319 U_CAPI int32_t U_EXPORT2
321  UChar *buf, int32_t cap, UErrorCode *status);
322 
323 
324 
353 U_CAPI UEnumeration * U_EXPORT2
355 
367 U_CAPI UBool U_EXPORT2
369 
370 
382 U_CAPI UBool U_EXPORT2
384 
385 #ifndef U_HIDE_INTERNAL_API
399 U_CAPI UEnumeration * U_EXPORT2
401 
415 U_CAPI void U_EXPORT2
416 ucsdet_setDetectableCharset(UCharsetDetector *ucsd, const char *encoding, UBool enabled, UErrorCode *status);
417 #endif /* U_HIDE_INTERNAL_API */
418 
419 #endif
420 #endif /* __UCSDET_H */
421 
422 
"Smart pointer" class, closes a UCharsetDetector via ucsdet_close().
C++ API: "Smart pointers" for use with and in ICU4C C++ code.
#define U_DEFINE_LOCAL_OPEN_POINTER(LocalPointerClassName, Type, closeFunction)
"Smart pointer" definition macro, deletes objects via the closeFunction.
Definition: localpointer.h:550
U_CAPI UBool ucsdet_enableInputFilter(UCharsetDetector *ucsd, UBool filter)
Enable filtering of input text.
U_CAPI int32_t ucsdet_getConfidence(const UCharsetMatch *ucsm, UErrorCode *status)
Get a confidence number for the quality of the match of the byte data with the charset.
struct UCharsetMatch UCharsetMatch
Opaque structure representing a match that was identified from a charset detection operation.
Definition: ucsdet.h:71
U_CAPI const UCharsetMatch ** ucsdet_detectAll(UCharsetDetector *ucsd, int32_t *matchesFound, UErrorCode *status)
Find all charset matches that appear to be consistent with the input, returning an array of results.
U_CAPI void ucsdet_close(UCharsetDetector *ucsd)
Close a charset detector.
U_CAPI UEnumeration * ucsdet_getDetectableCharsets(const UCharsetDetector *ucsd, UErrorCode *status)
Get an iterator over the set of detectable charsets - over the charsets that are enabled by the speci...
U_CAPI UBool ucsdet_isInputFilterEnabled(const UCharsetDetector *ucsd)
Test whether input filtering is enabled for this charset detector.
U_CAPI int32_t ucsdet_getUChars(const UCharsetMatch *ucsm, UChar *buf, int32_t cap, UErrorCode *status)
Get the entire input text as a UChar string, placing it into a caller-supplied buffer.
struct UCharsetDetector UCharsetDetector
Structure representing a charset detector.
Definition: ucsdet.h:63
U_CAPI void ucsdet_setDetectableCharset(UCharsetDetector *ucsd, const char *encoding, UBool enabled, UErrorCode *status)
Enable or disable individual charset encoding.
U_CAPI void ucsdet_setDeclaredEncoding(UCharsetDetector *ucsd, const char *encoding, int32_t length, UErrorCode *status)
Set the declared encoding for charset detection.
U_CAPI const char * ucsdet_getName(const UCharsetMatch *ucsm, UErrorCode *status)
Get the name of the charset represented by a UCharsetMatch.
U_CAPI UEnumeration * ucsdet_getAllDetectableCharsets(const UCharsetDetector *ucsd, UErrorCode *status)
Get an iterator over the set of all detectable charsets - over the charsets that are known to the cha...
U_CAPI const char * ucsdet_getLanguage(const UCharsetMatch *ucsm, UErrorCode *status)
Get the RFC 3066 code for the language of the input data.
U_CAPI const UCharsetMatch * ucsdet_detect(UCharsetDetector *ucsd, UErrorCode *status)
Return the charset that best matches the supplied input data.
U_CAPI void ucsdet_setText(UCharsetDetector *ucsd, const char *textIn, int32_t len, UErrorCode *status)
Set the input byte data whose charset is to detected.
U_CAPI UCharsetDetector * ucsdet_open(UErrorCode *status)
Open a charset detector.
C API: String Enumeration.
struct UEnumeration UEnumeration
structure representing an enumeration object instance
Definition: uenum.h:44
int8_t UBool
The ICU boolean type, a signed-byte integer.
Definition: umachine.h:269
#define U_CAPI
This is used to declare a function as a public ICU C API.
Definition: umachine.h:110
char16_t UChar
The base type for UTF-16 code units and pointers.
Definition: umachine.h:418
Basic definitions for ICU, for both C and C++ APIs.
UErrorCode
Standard ICU4C error code type, a substitute for exceptions.
Definition: utypes.h:415