Logo Search packages:      
Sourcecode: icu version File versions  Download package

ucnv_cb.h File Reference

C UConverter functions to aid the writers of callbacks. More...

#include "unicode/utypes.h"
#include "unicode/ucnv.h"
#include "unicode/ucnv_err.h"
Include dependency graph for ucnv_cb.h:
This graph shows which files directly or indirectly include this file:

Go to the source code of this file.


U_STABLE void U_EXPORT2 ucnv_cbFromUWriteBytes (UConverterFromUnicodeArgs *args, const char *source, int32_t length, int32_t offsetIndex, UErrorCode *err)
U_STABLE void U_EXPORT2 ucnv_cbFromUWriteSub (UConverterFromUnicodeArgs *args, int32_t offsetIndex, UErrorCode *err)
U_STABLE void U_EXPORT2 ucnv_cbFromUWriteUChars (UConverterFromUnicodeArgs *args, const UChar **source, const UChar *sourceLimit, int32_t offsetIndex, UErrorCode *err)
U_STABLE void U_EXPORT2 ucnv_cbToUWriteSub (UConverterToUnicodeArgs *args, int32_t offsetIndex, UErrorCode *err)
U_STABLE void U_EXPORT2 ucnv_cbToUWriteUChars (UConverterToUnicodeArgs *args, const UChar *source, int32_t length, int32_t offsetIndex, UErrorCode *err)

Detailed Description

C UConverter functions to aid the writers of callbacks.

Callback API for UConverter

These functions are provided here for the convenience of the callback writer. If you are just looking for callback functions to use, please see ucnv_err.h. DO NOT call these functions directly when you are working with converters, unless your code has been called as a callback via ucnv_setFromUCallback or ucnv_setToUCallback !!

A note about error codes and overflow. Unlike other ICU functions, these functions do not expect the error status to be U_ZERO_ERROR. Callbacks must be much more careful about their error codes. The error codes used here are in/out parameters, which should be passed back in the callback's error parameter.

For example, if you call ucnv_cbfromUWriteBytes to write data out to the output codepage, it may return U_BUFFER_OVERFLOW_ERROR if the data did not fit in the target. But this isn't a failing error, in fact, ucnv_cbfromUWriteBytes may be called AGAIN with the error status still U_BUFFER_OVERFLOW_ERROR to attempt to write further bytes, which will also go into the internal overflow buffers.

Concerning offsets, the 'offset' parameters here are relative to the start of SOURCE. For example, Suppose the string "ABCD" was being converted from Unicode into a codepage which doesn't have a mapping for 'B'. 'A' will be written out correctly, but The FromU Callback will be called on an unassigned character for 'B'. At this point, this is the state of the world: Target: A [..] [points after A] Source: A B [C] D [points to C - B has been consumed] 0 1 2 3 codePoint = "B" [the unassigned codepoint]

Now, suppose a callback wants to write the substitution character '?' to the target. It calls ucnv_cbFromUWriteBytes() to write the ?. It should pass ZERO as the offset, because the offset as far as the callback is concerned is relative to the SOURCE pointer [which points before 'C'.] If the callback goes into the args and consumes 'C' also, it would call FromUWriteBytes with an offset of 1 (and advance the source pointer).

Definition in file ucnv_cb.h.

Generated by  Doxygen 1.6.0   Back to index