Programming Forums > Application Development > C++

kurifu · Jul 7th, 2004, 3:47 AM

Alright, so I am writing a server client application that has to work cross platform, between Windows and linux. Now the deal is that the server and client can run on both platforms which makes managing the conversions a little more difficult.

To start off explaining this problem I am going to post a bit of output from some code I wrote to give me the sizes of a few various type on both linux and Windows.

(Toggle Plain Text)

Size of types in linux (gcc 2.3.3/gentoo 1.4rc1):
-------------------------------------------------

Sizeof: bool      1
Sizeof: int       4
Sizeof: unsigned int  4
Sizeof: long      4
Sizeof: double     8
Sizeof: float      4
Sizeof: unsigned char  1
Sizeof: char      1
Sizeof: wchar_t     4


Size of type in windows (MSVC++.net 2003), Windows XPsp2:
---------------------------------------------------------

Sizeof: bool      1
Sizeof: int       4
Sizeof: unsigned int  4
Sizeof: long      4
Sizeof: double     8
Sizeof: float      4
Sizeof: unsigned char  1
Sizeof: char      1
Sizeof: wchar_t     2

As you can see they are all pretty much the same, except wchar_t. On linux it is 4 bytes, I presume that it is using USC4 (if I recall correctly, I have this written down somewhere), however a quick googling has not really revealed to me what character encoding windows happens to be using... USC2? UTF16? And to make matter worse I have to figure out how to transfer all this data and preserve to data.

The most obvious, and probably best solution considering transfers will occure over a network is just to use the local setting and encode all the streams in UTF8 format before transferring since this tends to take less bandwidth when using latin based languages an characters and should provide an easy means for converting between most UCS and UTF formats.

Anyway, if anyone can tell me what format windows is using that is causing wchar_t to be represented by a 2 byte array I would greatly appreciate it, and maybe some info on converting between the two operating system.

Thanks....

kurifu · Jul 7th, 2004, 3:49 AM

Oh yeah, one thing I wanted to mention on top of that is that the 2byte representation of wchar_t is present when the application was compiled with MSVC.net 2003 (version 7) with MultiByte (MBCS) and UNICODE character sets enabled, the results were the same.

Jul 7th, 2004, 3:47 AM	#1
kurifu Expert Programmer Join Date: Jul 2004 Location: Halifax, Nova Scotia (Canada) Posts: 784 Rep Power: 5	Alright, so I am writing a server client application that has to work cross platform, between Windows and linux. Now the deal is that the server and client can run on both platforms which makes managing the conversions a little more difficult. To start off explaining this problem I am going to post a bit of output from some code I wrote to give me the sizes of a few various type on both linux and Windows. (Toggle Plain Text) Size of types in linux (gcc 2.3.3/gentoo 1.4rc1): ------------------------------------------------- Sizeof: bool 1 Sizeof: int 4 Sizeof: unsigned int 4 Sizeof: long 4 Sizeof: double 8 Sizeof: float 4 Sizeof: unsigned char 1 Sizeof: char 1 Sizeof: wchar_t 4 Size of type in windows (MSVC++.net 2003), Windows XPsp2: --------------------------------------------------------- Sizeof: bool 1 Sizeof: int 4 Sizeof: unsigned int 4 Sizeof: long 4 Sizeof: double 8 Sizeof: float 4 Sizeof: unsigned char 1 Sizeof: char 1 Sizeof: wchar_t 2 Size of types in linux (gcc 2.3.3/gentoo 1.4rc1): ------------------------------------------------- Sizeof: bool 1 Sizeof: int 4 Sizeof: unsigned int 4 Sizeof: long 4 Sizeof: double 8 Sizeof: float 4 Sizeof: unsigned char 1 Sizeof: char 1 Sizeof: wchar_t 4 Size of type in windows (MSVC++.net 2003), Windows XPsp2: --------------------------------------------------------- Sizeof: bool 1 Sizeof: int 4 Sizeof: unsigned int 4 Sizeof: long 4 Sizeof: double 8 Sizeof: float 4 Sizeof: unsigned char 1 Sizeof: char 1 Sizeof: wchar_t 2 As you can see they are all pretty much the same, except wchar_t. On linux it is 4 bytes, I presume that it is using USC4 (if I recall correctly, I have this written down somewhere), however a quick googling has not really revealed to me what character encoding windows happens to be using... USC2? UTF16? And to make matter worse I have to figure out how to transfer all this data and preserve to data. The most obvious, and probably best solution considering transfers will occure over a network is just to use the local setting and encode all the streams in UTF8 format before transferring since this tends to take less bandwidth when using latin based languages an characters and should provide an easy means for converting between most UCS and UTF formats. Anyway, if anyone can tell me what format windows is using that is causing wchar_t to be represented by a 2 byte array I would greatly appreciate it, and maybe some info on converting between the two operating system. Thanks.... __________________ Clifford Matthew Roche <geek@cliffordroche.com> Web Hosting: http://www.crd-hosting.com Consulting: http://www.crdev-consulting.com

Jul 7th, 2004, 3:49 AM	#2
kurifu Expert Programmer Join Date: Jul 2004 Location: Halifax, Nova Scotia (Canada) Posts: 784 Rep Power: 5	Oh yeah, one thing I wanted to mention on top of that is that the 2byte representation of wchar_t is present when the application was compiled with MSVC.net 2003 (version 7) with MultiByte (MBCS) and UNICODE character sets enabled, the results were the same. __________________ Clifford Matthew Roche <geek@cliffordroche.com> Web Hosting: http://www.crd-hosting.com Consulting: http://www.crdev-consulting.com

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Programming Forums > Application Development > C++

A bit of UNICODE anyone?