Need help with longs in file opening (SOLVED)

[Gregjazz]

I'm working on a project right now that requires to be able to read a file format. So far so good, except in one section of the file format it uses longs.

I actually can't find much about longs in the AGS help documentation.

I have the file format specs, so I know how it works, but it's a little difficult interpreting it into AGS. In the part of the file format I'm working on now, the file has several parameters supposedly stored as longs. Up until then, it's used "chars", and I know how to work those in AGS.

Please any details on how this works would be appreciated. How do you interpret a file format in AGS?

[Pumaman]

What type of "long" are they? If it's a 32-bit long, then that's the equivalent of an AGS "int".

[Gregjazz]

Well actually I'm probably getting the terminology wrong.

I have a .cpp that says the file format works like this:


fwrite("chnk",4,1,verbose);
fwrite(&Waveforms.Wave

• .Size,      sizeof(long), 1,    verbose);
  fwrite(&Waveforms.Wave
• .VelocityLow,   sizeof(long), 1,    verbose);
  fwrite(&Waveforms.Wave
• .VelocityHigh,   sizeof(long), 1,    verbose);
  fwrite(&Waveforms.Wave
• .KeyLow,   sizeof(long), 1,    verbose);
  fwrite(&Waveforms.Wave
• .KeyHigh,   sizeof(long), 1,    verbose);
  fwrite(&Waveforms.Wave
• .loopStart,   sizeof(long), 1,    verbose);
  fwrite(&Waveforms.Wave
• .LoopEnd,   sizeof(long), 1,    verbose);
  fwrite(&Waveforms.Wave
• .Root,      sizeof(long), 1,    verbose);
  fwrite(&Waveforms.Wave
• .Samplerate,   sizeof(long), 1,    verbose);
  fwrite(&Waveforms.Wave
• .Channel,   sizeof(long), 1,    verbose);
  [/i]
  
  And in hexedit the example file at that place goes like this:
  
  

  Code: ags

  
                                                                 63 68 6E 6B
  DF 00 00 00 00 00 00 00    7F 00 00 00 16 00 00 00
  6B 00 00 00 FF FF FF FF      00 00 00 00 18 00 00 00
  44 AC 00 00 00 00 00 00
  

  
  
  So "63 68 6E 6B" is the "chnk" part.
  
  "DF 00 00 00" is the waveform size I guess. And so forth. As you can see the data is in groups of 4 numbers. I'm not sure how to interpret this in AGS...
  
  I know it's a whole lot to ask, but would anyone be willing to give me some tips?

[HeirOfNorton]

That is a 32-bit integer, the same type as used in AGS.  Each of those four numbers output by a hex-editor is a single byte, with four bytes (32 bits) per variable. I could mention endianness as well, but from the looks of those numbers it should not be necessary. Just use File.ReadRawInt and it should read them correctly.

HoN

[Gregjazz]

That is a 32-bit integer, the same type as used in AGS.  Each of those four numbers output by a hex-editor is a single byte, with four bytes (32 bits) per variable. I could mention endianness as well, but from the looks of those numbers it should not be necessary. Just use File.ReadRawInt and it should read them correctly.

HoN

Well it makes sense, but all it returns are -1's, using the File.ReadRawInt.

[Kweepa]

Post your code here.

It looks (from the hex dump) like the 'CHNK' marker is not the first thing in the file, so presumably you're searching for that, then reading the subsequent numbers into a Wave structure.
I would expect loopStart to be -1 and the others, not.

[Gregjazz]

Post your code here.

It looks (from the hex dump) like the 'CHNK' marker is not the first thing in the file, so presumably you're searching for that, then reading the subsequent numbers into a Wave structure.
I would expect loopStart to be -1 and the others, not.

Exactly.

Here's the code:

Code: ags

File *output = File.Open("Sampleset.WusikSND", eFileRead);
if (output != null) {
  int p;
  int pa;
  while ((pa != 1) || (output.EOF == false)) {
    int temp = output.ReadRawChar();
    string buffer;
    StrFormat(buffer,"%c", temp);
    if ((StrComp(buffer,"c") == 0) && (p == 0)) {
      p += 1;
      }
    else if ((StrComp(buffer,"h") == 0) && (p == 1)) {
      p += 1;
      }
    else if ((StrComp(buffer,"n") == 0) && (p == 2)) {
      p += 1;
      }
    else if ((StrComp(buffer,"k") == 0) && (p == 3)) {
      pa = 1;
      }
    else {
      p = 0;
      }
    }
    
  p = 0;
  while (p < 5) {
     int data;
     data = output.ReadRawInt();
     Display("%d", data);
     p += 1;
     }
  output.Close();
  }

Really not the best way of doing things, but it searches for the "CHNK" and then reads the subsequent data.

[Kweepa]

That looks like it should almost work.
The problem is most likely the || in the while statement - it should be && (the way you have it, the while loop always reads to the end of the file, and then the data reading reads -1s from "past end of the file".
There are a couple of things that I'd do differently to make sure it works as expected.
Firstly, initialize your variables. I'm pretty sure AGS initializes them to zero anyway, but it can't hurt.
Secondly, you can search for the whole 'chnk' piece at once. That way you don't accidentally miss a "find" when 'c' appears, then 'h', then 'c' 'h' 'n' 'k'. Or you could just reset the search to 1 when 'c' comes up.
Lastly, I renamed output to input

[EDIT] The fourByteBoundaryOk "variable" is separating the two approaches to finding 'chnk'. The first, simpler approach, assumes that 'chnk' comes on a 4 byte boundary in the file, hence the "variable" name. For safety's sake, probably better to use the second approach.

Code: ags

File *input = File.Open("Sampleset.WusikSND", eFileRead);
if (input != null)
{
Ã,  bool foundChnk = false;
Ã,  int chnkStage = 0;
Ã,  while (!foundChnk && !input.EOF)
Ã,  {
Ã,  Ã,  if (fourByteBoundaryOk)
Ã,  Ã,  {
Ã,  Ã,  Ã,  int candidate = input.ReadRawInt();
Ã, Ã,  Ã, Ã,  if (candidate == 1802397795) // 0x6b6e6863 in decimal
Ã,  Ã, Ã,  Ã, {
Ã,  Ã,  Ã, Ã,  Ã, foundChnk = true;
Ã,  Ã, Ã,  Ã, }
Ã,  Ã,  }
Ã,  Ã,  else
Ã,  Ã,  {
Ã,  Ã,  Ã,  // put the values into a string
Ã,  Ã,  Ã,  string chnk = "chnk";
Ã,  Ã,  Ã,  int candidate = input.ReadRawChar();
Ã,  Ã,  Ã,  if (candidate == StrGetCharAt(chnk, chnkStage))
Ã,  Ã,  Ã,  {
Ã,  Ã,  Ã,  Ã,  // note - this only works for finding strings with no repeating patterns
Ã,  Ã,  Ã,  Ã,  chnkStage++;
Ã,  Ã,  Ã,  Ã,  if (chnkStage == 4) chnkFound = true;
Ã,  Ã,  Ã,  }
Ã,  Ã,  Ã,  else
Ã,  Ã,  Ã,  if (candidate == StrGetCharAt(chnk, 0))
Ã,  Ã,  Ã,  {
Ã,  Ã,  Ã,  Ã,  // reset
Ã,  Ã,  Ã,  Ã,  chnkStage = 1;
Ã,  Ã,  Ã,  }
Ã,  Ã,  Ã,  else
Ã,  Ã,  Ã,  {
Ã,  Ã,  Ã,  Ã,  chnkStage = 0;
Ã,  Ã,  Ã,  }
Ã,  Ã,  }
Ã,  }
Ã,  if (foundChnk)
Ã,  {
Ã,  Ã,  int size = input.ReadRawInt();
Ã,  Ã,  int velocityLow = input.ReadRawInt();
Ã,  Ã,  int velocityHigh = input.ReadRawInt();
Ã,  Ã,  // etc
Ã,  Ã,  Display("size = %d, velocityLow = %d, velocityHigh = %d", size, velocityLow, velocityHigh);
Ã,  }
Ã,  else
Ã,  {
Ã,  Ã,  Display("Didn't find chnk. Perhaps I got chnk's decimal value wrong.");
Ã,  }
Ã,  input.Close();
}
else
{
Ã,  Display("Couldn't open file. Bad user! Bad!");
}

[Gregjazz]

What is the "if (fourByteBoundaryOk)" part?

[Gregjazz]

It works brilliantly now!!

I'm going to study the code to learn from it now.

Thanks so much for your help.