Created
June 12, 2011 19:35
-
-
Save m039/1021913 to your computer and use it in GitHub Desktop.
Print a content of an utf16 file to the stdout.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
/* Author: m039 <flam44 (at) gmail (dot) com> */ | |
#include <stdio.h> | |
/* | |
* Use the next command to test this snippet: | |
* | |
* iconv -f utf8 -t utf16 utf8.txt -o utf16.txt | |
* ./utf16_to_utf8 utf16.txt > utf8.txt | |
*/ | |
static FILE *g_fp; | |
static char g_buffer[512]; /* The buffer is too small for big files */ | |
typedef unsigned int character; | |
void utf8(character ch, char **string) { | |
char *src; | |
if (*string == NULL) { | |
return; | |
} | |
src = *string; | |
if (ch <= 0x7f) { | |
src[0] = ch; | |
*string += 1; | |
} else if (ch <= 0x0007ff) { | |
src[0] = 0xc0 | (ch >> 6) & 0x1f; | |
src[1] = 0x80 | (ch & 0x3f); | |
*string += 2; | |
} else if (ch <= 0x00ffff) { | |
src[0] = 0xe0 | (ch >> 12) & 0x0f; | |
src[1] = 0x80 | (ch >> 6) & 0x3f; | |
src[2] = 0x80 | (ch & 0x3f); | |
*string += 3; | |
} else if (ch <= 0x10ffff) { | |
src[0] = 0xf0 | (ch >> 18) & 0x07; | |
src[1] = 0x80 | (ch >> 12) & 0x3f; | |
src[2] = 0x80 | (ch >> 6) & 0x3f; | |
src[3] = 0x80 | (ch & 0x3f); | |
*string += 4; | |
} | |
} | |
int fgetshort(FILE *fp) { | |
int c1, c2; | |
c1 = fgetc(fp); | |
c2 = fgetc(fp); | |
if ((c1 == EOF) || (c2 == EOF)) { | |
return EOF; | |
} | |
return ((c2 << 8) | c1) & 0xffff; | |
} | |
int main(int argc, char **argv) { | |
int c, surrogate; | |
char *p = g_buffer; | |
if (argc < 2) { | |
fprintf(stderr, "Usage: ./utf16_to_utf8 <from>\n"); | |
return -1; | |
} | |
g_fp = fopen(argv[1], "r"); | |
if (g_fp == NULL) { | |
perror(argv[1]); | |
return -1; | |
} | |
while ((c = fgetshort(g_fp)) != EOF) { | |
if ((c >= 0xd800) && (c <= 0xdbff)) { | |
surrogate = (c - 0xd800) << 10; | |
} else if ((c >= 0xdc00) && (c <= 0xdfff)) { | |
surrogate |= c - 0xdc00; | |
utf8(surrogate + 0x10000, &p); | |
} else { | |
utf8(c, &p); | |
} | |
} | |
*p = 0; | |
fclose(g_fp); | |
puts(g_buffer); | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment