Re: [xml] TestC14N patch (standard C output)

[Justin Fletcher <justin fletcher ntlworld com>]

In message <3C8AC51F 3000809 aleksey com>
          Aleksey Sanin <aleksey aleksey com> wrote:

> Then it probably also a good idea to remove
>     #include <unistd.h>
> from this file. I do not remember exactly why I used write() instead of
> fwrite() here. Probably because I recently had problems with Solaris
> struct _FILE because it is declared as follows:
>
> struct _FILE {
>     ....
>     unsinged char fd;
>     ....
> };
>
> and this effectevly limits the number of files opened with FILE* fopen() to

The "effective limit" on the number of open FILE * is FOPEN_MAX (7.19.1.3
introduction to 'Input/output <stdio.h>') which is declared to be the
minimum number of file that the implementation guarantees can be open
simultaneously. This shall be at least 8 (7.19.4.2.15; environmental
limits).

> 256 per process.  After this I am trying to avoid any FILE* usage as much
> as possible. Of course, there is nothing to do with this particular case,
> just automatic reaction to f<something> :)

Ah. I'm quite the opposite; Every single read/write/open/close call sets me
swearing. They're not standard C and therefore not available to the compiler
I use, thus I have to remove them at every turn. From a platform
independence point of view, non-standard C IO functions make my life hell.

[snip]

-- 
Gerph {djf0-.3w6e2w2.226,6q6w2q2,2.3,2m4}
URL: http://www.movspclr.co.uk/
... Eyes to the heavens, screaming at the sky;
    Trying to send you messages, but choking on goodbye.