Parsing parameters from a text file in a random order

Good afternoon,

I am trying to write a module to read an unknown, albeit very small (10-15), number of parameters from a text file( such as "Width: 200\n" ). However, I want to be able to specify the parameters in any order and omit some if necessary.

I figured I should loop through all the lines using fscanf, continuously checking the return value after each iteration and acting appropriately. (i.e if the parameter is read, exit the loop and carry on to the next parameter, otherwise if we hit the EOF marker and the value was not found, assign it a
default value)

Code:

width = -1; // assign a dummy value

while ( (fc = fscanf(filePtr, "Width:%d\n", &width)) != EOF )
{
   if( fc == 1)
      break; // parameter was read so exit the loop
}

if( width == -1 )
   width = 300; // parameter not found so default it

However, this approach doesn't work for some reason.
Any help would be appreciated.

Duetti

How about trying the old fgets/sscanf combo?

Code:

char buffer[BUFSIZ];
int width = -1;

while ( fgets(buffer, sizeof buffer, filePtr) )
{
   if ( sscanf(buffer, "Width:%d", &width) == 1 )
   {
      break;
   }
}

if ( width == -1 )
{
   width = 300;
}

Thanks for the tip Dave. This combo works fine indeed.

Code:

while ( fgets(buffer, 1024, in) )
{
   // the first condition will fail so the second will be discarded automatically
   if ( widthRead == 0 && sscanf(buffer, "Width:%d",  &gContext.width) == 1 )
      widthRead = !widthRead;
   else if ( heightRead == 0 && sscanf(buffer, "Height:%d", &gContext.height) == 1 )
      heightRead = 1;
   else if ( initXYRead == 0 && sscanf(buffer, "InitXY(%d,%d)", &gContext.initX, &gContext.initY) == 2 )
      initXYRead = 1;
   else if ( CColRead == 0 && sscanf(buffer, "CCol:(%f,%f,%f)", &gContext.CCol.r, &gContext.CCol.g, &gContext.CCol.b) == 3 )
      CColRead = 1;
   else if ( i<10 && sscanf(buffer, "RGB(%f,%f,%f)", &gContext.preCol[i].r, &gContext.preCol[i].g, &gContext.preCol[i].b) == 3 )
      i++;}
}

I wonder why the other piece of code didn't work...
This solution does what it was designed to do, however it seems like an overkill to me (even though I can't think of a more elegant solution).

Duetti

> I wonder why the other piece of code didn't work...
Because you blew your chance to try and reparse the data with a different conversion (you used fscanf).

With fgets(), you have as many goes as it takes with the SAME buffer to decide what it all means

>Because you blew your chance to try and reparse the data with a different conversion (you used fscanf).

What do you mean?

use fscanf twice? once in the loop continuation condition and the second time inside the body of the loop?

Duetti

fscanf() reads the file, and tries to parse the data.
The problem is, you've already read the file, so there's no real possibility of backing up and trying again with a different conversion string.

But I used rewind() to go back to the beginning of the file, i just didn't include the function call in the above code fragment.

The problem with the initial code fragment was that it couldn't locate the string "Width:%d" anywhere in the file, even though it was there.

>>even though it was there.
Using fscanf() means that if the data read doesn't match your requirements, the read stops. If you have a line of text your program isn't expecting at all, you'll find yourself in an endless loop.

The best way, as shown, is to use fgets() then parse each line in memory.

Yes, the read stops but the file pointer now points to the next line, right? Will it not keep reading lines until the EOF is reached or the string is found? If you test for the EOF after each iteration, how can you get stuck in an infinite loop?

>Yes, the read stops but the file pointer now points to the next line, right?

No. Reading by lines is done with fgets. [edit]Or very thorough and careful understanding of how to use fscanf's directives.[/edit]

>Will it not keep reading lines until the EOF is reached or the string is found?

No.