🚀 go-pugleaf

In article <199511272135.WAA02816@rulcmc.LeidenUniv.nl> joost@rulcmc.leidenuniv.nl (joost witteveen) writes:
>From: joost@rulcmc.leidenuniv.nl (joost witteveen)
>Subject: initialization of ostrstream
>Date: 27 Nov 1995 16:37:47 -0500

>If I construct a ostrstream, with
>  ostrstream os (s,sizeof(s))
>do I have to clean ALL of s before I use it?
>I'm getting strange results (i.e. I have to init ALL of s to 0)
>with g++ 2.7.1, and 2.3.3.

>So, my question is: is this a bug in g++, or am I missing something
>about the string streams?


>The following programme has this what I think is strange behaviour,
>as it outputs in the last line not only the "hello, world", but
>part of the old h's string s was inititalized to,  concatenated to it.

>Thanks for any suggestions.


>#include <iostream.h>
>#include <strstream.h>
>int main(int,char **){
>  int i;
>  char s[]="\000hhhhhhhhhhhhhhhhhhhhhhhhhhhh";
>  cout<<"initial s:"<<s<<endl; //an empty string here.

>  ostrstream os(s,sizeof(s));
>  os<<"hello, world";
>  cout<<s<<endl;          //outputs: hello, worldhhhhhhhhhhhhhhhhh
>}

>-- 

>joost witteveen
>            joost@rulcmc.leidenuniv.nl
>          joostje@dds.hacktic.nl


You are missing an ends (insertor for the string terminator)



Kit Kauffmann - kitk@mudshark.sunquest.com
AKA 73363,447 (Compu$erve)

Finger me for my public key

In article <199511272135.WAA02816@rulcmc.LeidenUniv.nl> joost@rulcmc.leidenuniv.nl (joost witteveen) writes:

   If I construct a ostrstream, with
     ostrstream os (s,sizeof(s))
   do I have to clean ALL of s before I use it?
   I'm getting strange results (i.e. I have to init ALL of s to 0)
   with g++ 2.7.1, and 2.3.3.

   So, my question is: is this a bug in g++, or am I missing something
   about the string streams?


   The following programme has this what I think is strange behaviour,
   as it outputs in the last line not only the "hello, world", but
   part of the old h's string s was inititalized to,  concatenated to it.

   Thanks for any suggestions.


   #include <iostream.h>
   #include <strstream.h>
   int main(int,char **){
     int i;
     char s[]="\000hhhhhhhhhhhhhhhhhhhhhhhhhhhh";
     cout<<"initial s:"<<s<<endl; //an empty string here.

     ostrstream os(s,sizeof(s));
     os<<"hello, world";
     cout<<s<<endl;          //outputs: hello, worldhhhhhhhhhhhhhhhhh
   }

You've missed to terminate the string. Use 'ends' for an additional
0-char at the end of the string, ie. exchange 'os<<"hello, world";'
by 'os<<"hello, world"<<ends;'.

	regards
		Enno

joost witteveen (joost@rulcmc.leidenuniv.nl) wrote:
>  If I construct a ostrstream, with
>    ostrstream os (s,sizeof(s))
>  do I have to clean ALL of s before I use it?
>  I'm getting strange results (i.e. I have to init ALL of s to 0)
>  with g++ 2.7.1, and 2.3.3.

>  So, my question is: is this a bug in g++, or am I missing something
>  about the string streams?


>  The following programme has this what I think is strange behaviour,
>  as it outputs in the last line not only the "hello, world", but
>  part of the old h's string s was inititalized to,  concatenated to it.

>  Thanks for any suggestions.


>  #include <iostream.h>
>  #include <strstream.h>
>  int main(int,char **){
>    int i;
>    char s[]="\000hhhhhhhhhhhhhhhhhhhhhhhhhhhh";
>    cout<<"initial s:"<<s<<endl; //an empty string here.

>    ostrstream os(s,sizeof(s));
>    os<<"hello, world";
>    cout<<s<<endl;          //outputs: hello, worldhhhhhhhhhhhhhhhhh
>  }

You forgot to append a '\0', which is _not_ done automatically by
'ostrstream'. There is a special manipulator 'ends' which does exactly
this:

   os << "hello world" << ends;

Now the string is correctly '\0'-terminated.

	Bernd

--
+----------------------------------+
|          Bernd Eggink            |
|    Rechenzentrum Uni Hamburg     |
| admin@rzaix13.rrz.uni-hamburg.de |
+----------------------------------+