Intravision/reactos
1
0
Fork 0
mirror of https://github.com/reactos/reactos.git synced 2025-02-23 08:55:19 +00:00
Code Issues Projects Releases Packages Wiki Activity

[UTF16LE]

Browse source 
- Implement ability to disable BOM writing

svn path=/trunk/; revision=72426
This commit is contained in:
Dmitry Chapyshev 2016-08-22 13:14:41 +00:00
parent 6ba5e47c0e
commit 457d4cff98
1 changed files with 30 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

38
reactos/sdk/tools/utf16le/utf16le.cpp
View file

@ -13,6 +13,7 @@
#include <fstream>
#include <iostream>
#include <string.h>
//#define DISPLAY_DETECTED_UNICODE
@ -25,14 +26,16 @@ public:
// due to ambiguous BOM
enum enc_types { detect, utf8, utf16le, utf16be, utf32le, utf32be };
enum err_types { none, iopen, oopen, eof, read, write, decode };
enum bom_types { bom, nobom };
protected:
err_types error;
enc_types encoding;
bom_types bom_type;
unsigned char buffer[4], fill, index; // need 4 char buffer for optional BOM handling
fstream inputfile,outputfile;
static const unsigned char utf8table[64];
public:
utf_converter(string ifname, string ofname, enc_types enc = detect) : error(none), encoding(enc), fill(0), index(0)
utf_converter(string ifname, string ofname, bom_types ofbom = bom, enc_types enc = detect) : error(none), bom_type(ofbom), encoding(enc), fill(0), index(0)
{
enc_types tmp_enc;
inputfile.open(ifname.c_str(), ios::in | ios::binary);
@ -230,10 +233,15 @@ public:
}
void convert2utf16le()
{
wchar_t c;
unsigned char buffer[2] = {0xff, 0xfe};
outputfile.write(reinterpret_cast<char*>(&buffer),2); // write BOM
c = get_wchar_t();
unsigned char buffer[2] = { 0xff, 0xfe };
if (bom_type == bom)
{
outputfile.write(reinterpret_cast<char*>(&buffer), 2); // write BOM
}
wchar_t c = get_wchar_t();
while (!inputfile.eof())
{
buffer[0] = c & 0xff;
@ -262,12 +270,22 @@ const unsigned char utf_converter::utf8table[64] = {
int main(int argc, char* argv[])
{
utf_converter::err_types err;
if (argc < 3)
{
cout << "usage: " << argv[0] << " inputfile outputfile" << endl;
return -1;
}
utf_converter conv(argv[1],argv[2]);
utf_converter::bom_types bom_type = utf_converter::bom;
if (argc == 4 && _stricmp(argv[3], "nobom") == 0)
{
bom_type = utf_converter::nobom;
}
utf_converter conv(argv[1],argv[2], bom_type);
if ((err = conv.getError())!=utf_converter::none)
{
switch (err)
@ -282,7 +300,11 @@ int main(int argc, char* argv[])
cerr << "Unknown error." << endl;
}
return -1;
} else
conv.convert2utf16le();
}
else
{
conv.convert2utf16le();
}
return 0;
}