Integer.parseInt("0") fails

Hi,
I'm parsing the reply from a webside. Parsing the header and splitting header and body works, but then it fails.
The Webside itself just shows things like:

Code:

0
1 bla
2
3 foo

and that's the parsing code:

Code:java

String[] split;
if(body.contains("\n"))
 split = body.split("\n");
else
 split = new String[] { body };
int[] cmds = new int[split.length];
try
{
 for(int i = 0; i < split.length; i++)
 {
if(split[i].equals(""))
 continue;
System.out.print("Checking: '"+split[i]+"'");
if(!split[i].contains(" "))
{
 cmds[i] = Integer.parseInt(split[i]);
 split[i] = null;
}
else
{
 sp = split[i].indexOf(" ");
 cmds[i] = Integer.parseInt(split[i].substring(0, sp));
 split[i] = split[i].substring(sp + 1);
}
 }
}
catch (NumberFormatException e)
{
  System.out.print("Check: Invalid command");
 return;
}[/i][/i][/i][/i][/i][/i][/i][/i][/i][/i][/i]

In the logs I get:

Code:

Checking: '0'
Check: Invalid command

 

Do you maybe have a \r in there? Try printing

Code:

split[i].length()

 

I was thinking the same, maybe you have an \r\n, \n or \r in there.

 

Not possible as I filter out all \r:
reply = reply.replaceAll("\r", "");
and split by \n:

Code:java

 
if(body.contains("\n"))
split = body.split("\n");
else
split = new String[] { body };

But the length test goes in the right direction: I extended the debugging line:

Code:java

System.out.print("Checking: '"+split[I]+"': "+split[I].length());[/I][/I]

And the output:
Checking: '0': 776
Why the hell 776?!


//EDIT: Got it! The bad character was \0

 

Why don't you simply use split = body.split("\n")?