以前我寫了一個gb18030到utf-8編碼轉換的程序,這段代碼還是有些問題的,因爲現在我需要對任意的兩個iconv支持的語言編碼做互相轉換,比如GB2312, GBK, GB18030, UTF-8, UTF-16, BIG5等等,所以纔有了這段程序,註釋我不加了。目前這段代碼是非常的穩定,測試了超過10萬行的數十種編碼的文本的轉換都沒有出問題。
#include <stdlib.h>
#include <string.h>
#include <iconv.h>
#ifndef ICONV_CONST
# define ICONV_CONST const
#endif
/*!
對字符串進行語言編碼轉換
param from 原始編碼,比如"GB2312",的按照iconv支持的寫
param to 轉換的目的編碼
param save 轉換後的數據保存到這個指針裏,需要在外部分配內存
param savelen 存儲轉換後數據的內存大小
param src 原始需要轉換的字符串
param srclen 原始字符串長度
*/
int
convert(const char *from, const char *to, char* save, int savelen, char *src, int srclen)
{
iconv_t cd;
char *inbuf = src;
char *outbuf = save;
size_t outbufsize = savelen;
int status = 0;
size_t savesize = 0;
size_t inbufsize = srclen;
const char* inptr = inbuf;
size_t insize = inbufsize;
char* outptr = outbuf;
size_t outsize = outbufsize;
cd = iconv_open(to, from);
iconv(cd,NULL,NULL,NULL,NULL);
if (inbufsize == 0) {
status = -1;
goto done;
}
while (insize > 0) {
size_t res = iconv(cd,(ICONV_CONST char**)&inptr,&insize,&outptr,&outsize);
if (outptr != outbuf) {
int saved_errno = errno;
int outsize = outptr - outbuf;
strncpy(save+savesize, outbuf, outsize);
errno = saved_errno;
}
if (res == (size_t)(-1)) {
if (errno == EILSEQ) {
int one = 1;
iconvctl(cd,ICONV_SET_DISCARD_ILSEQ,&one);
status = -3;
} else if (errno == EINVAL) {
if (inbufsize == 0) {
status = -4;
goto done;
} else {
break;
}
} else if (errno == E2BIG) {
status = -5;
goto done;
} else {
status = -6;
goto done;
}
}
}
status = strlen(save);
done:
iconv_close(cd);
return status;
}