DaniWeb Forum Index > Software Development > C++

C++ RSS

The difference of '\n' in windows and linux.

•

Join Date: Jul 2006

Posts: 647

Reputation: jaepi is an unknown quantity at this point

Solved Threads: 4

jaepi

Offline

Practically a Master Poster

The difference of '\n' in windows and linux.

Nov 9th, 2007

Hello there. Is there anyway that you can convert the '\n' in windows which is 2bytes to the '\n' 1byte so that the linux could read it?

Retreat!!!

•

Join Date: Jul 2006

Posts: 647

Reputation: jaepi is an unknown quantity at this point

Solved Threads: 4

jaepi

Offline

Practically a Master Poster

Re: The difference of '\n' in windows and linux.

Nov 9th, 2007

In addition, does the fread() function can use wchar_t* type as buffer?

Retreat!!!

•

Join Date: Oct 2007

Posts: 1,951

Reputation: Duoas has much to be proud of

Solved Threads: 214

Duoas

Offline

Posting Virtuoso

Re: The difference of '\n' in windows and linux.

Nov 9th, 2007

•

Is there anyway that you can convert the '\n' in windows which is 2bytes to the '\n' 1byte so that the linux could read it?

Sure. There is a little Unix utility called dos2unix. It is found on most *nix systems.

To do it in code, just set your output file mode to "binary". Doing so turns off the '\n' to "\r\n" translations, so when you write your file you get exactly what you write --single newline and all.

•

In addition, does the fread() function can use wchar_t* type as buffer?

No, it uses char. Sorry.

However, there are usually functions to convert a binary string to a unicode string...
(I don't know what they are though.)

Hope this helps.

•

Join Date: Dec 2006

Posts: 1,089

Reputation: vijayan121 is a name known to all

Solved Threads: 164

vijayan121 vijayan121 is offline

Offline

Veteran Poster

Re: The difference of '\n' in windows and linux.

Nov 9th, 2007

converting ascii text files is trivial; you have a variety of tools. eg. tr, perl, awk, vi etc.in fact anything that can replace a cr-lf sequence with just a cr. eg.
awk '{ sub("\r$", ""); print }' winfile.txt > unixfile.txt
you could also use a utility called tofrodos which is available on a number of systems (under different names like "todos", "fromdos", "dos2unix", "unix2dos" etc.)
http://kb.iu.edu/data/acux.html
http://www.thefreecountry.com/tofrodos/

if the files are in unicode, these simple techniques will not work. unicode has a variety of newlines: CR (000D), LF (000A), CRLF (000D,000A), NEL (0085), FF (000C), LS (2028), and PS (2029). for a perl hack see http://www.onlamp.com/pub/a/onlamp/2...es.html?page=3. it is easy in c++ to write a function that processes input character by character and does the necessary translations.

fread and fwrite are to i/o what memcpy is to memory. it reads or writes bytes into untyped memory (void*) without any formatting or interpretation. the stream should be opened in binary ("b") mode to prevent newline character translations by the underlying layers. for example,

 Help with Code Tags 
 C++ Syntax (Toggle Plain Text) 
int array[100] ;
// to write an array (binary) into afile 
fwrite( array, sizeof(int), 100, file ); 
// to read the array (binary) from a file 
fread( array, sizeof(int), 100, file );
 
struct some_struct mystruct  ;
// write
fwrite( &mystruct, sizeof(struct some_struct), 1, file ) ; 
// read
fread( &mystruct, sizeof(struct some_struct), 1, file ) ;
int array[100] ;
// to write an array (binary) into afile 
fwrite( array, sizeof(int), 100, file ); 
// to read the array (binary) from a file 
fread( array, sizeof(int), 100, file );

struct some_struct mystruct  ;
// write
fwrite( &mystruct, sizeof(struct some_struct), 1, file ) ; 
// read
fread( &mystruct, sizeof(struct some_struct), 1, file ) ;

using these functions correctly is difficult; mainly because they do not distinguish between end of file and error. use of feof and ferror is required to determine what has occurred.