fread misreading character 255 1

[robherc]

Hi; I've worked out all but 1 of the bugs in my first truly functional program in C.

The Bug:
Every time the fread() function in my program comes across a 0xFF character in the source file, it's feeding *something* outside the bounds of 0-255 into the corresponding unsigned char in my buffer array.

The fread call:

readLength = fread(inBytes, 1, 1000, inFile);

inBytes (defined earlier as unsigned char inBytes[1000]) is then passed into a switch:case function that tests each element of the array for values 0-255 inclusive. Characters are then converted to hexadecimal values & output into another file.
I can directly feed values 0-255 into elements of the inBytes array, and the output is perfect (no missed characters), but when fed from a file opened with:

inFile = fopen(fileName, "rb");

The output was missing several characters.
Noticing the error in the output, I added a default: case to the switch that inserted '--' in any "uncaught" character; then my output file had proper spacing, and I was able to compare the output to a different (not written by me) conversion program. From the output compares, I found that only "FF" characters were affected. Each "FF" character put out from the other program corresponded exactly to an "--" from my program; so what it causing this non-numeric output from fread() for only that one character?

Obviously, I could alter my default: case to output "FF" instead of "--", but that doesn't fix the *cause* of my glitch, so I can't be certain it won't lead to other conversion errors in the future if I do it; much better to find, understand, then fix the original glitch, than to apply a sloppy workaround.

BTW: I'm compiling using gcc on an x86_64 computer running 64-bit OpenSuse Linux, if it matters (although I'm trying to write the program to be fully portable to other platforms).

 

[Salem]

How sure are you of the contents of the file (have you looked with a hex editor?)

Can you adapt this runnable test case to make it show your problem?

#include<stdio.h>
int main (void)
{
  FILE *fp = fopen("test.bin","wb");
  for ( int i = 0 ; i <= 255 ; i++ ) {
    fputc(i,fp);
  }
  fclose(fp);
  fp = fopen("test.bin","rb");
  unsigned char buff[256];
  fread(buff,sizeof(buff),1,fp);
  fclose(fp);
  for ( int i = 0 ; i < 256 ; i++ ) {
    switch ( buff[i] ) {
      case 0:
        printf("Found 0 at position %d\n", i );
        break;
      case 255:
        printf("Found 255 at position %d\n", i );
        break;
    }
  }
  return 0;
}

$ gcc -std=c99 foo.c
$ ./a.out 
Found 0 at position 0
Found 255 at position 255
$ odx test.bin 
000000 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f  >................<
000010 10 11 12 13 14 15 16 17 18 19 1a 1b 1c 1d 1e 1f  >................<
000020 20 21 22 23 24 25 26 27 28 29 2a 2b 2c 2d 2e 2f  > !"#$%&'()*+,-./<
000030 30 31 32 33 34 35 36 37 38 39 3a 3b 3c 3d 3e 3f  >0123456789:;<=>?<
000040 40 41 42 43 44 45 46 47 48 49 4a 4b 4c 4d 4e 4f  >@ABCDEFGHIJKLMNO<
000050 50 51 52 53 54 55 56 57 58 59 5a 5b 5c 5d 5e 5f  >PQRSTUVWXYZ[\]^_<
000060 60 61 62 63 64 65 66 67 68 69 6a 6b 6c 6d 6e 6f  >`abcdefghijklmno<
000070 70 71 72 73 74 75 76 77 78 79 7a 7b 7c 7d 7e 7f  >pqrstuvwxyz{|}~.<
000080 80 81 82 83 84 85 86 87 88 89 8a 8b 8c 8d 8e 8f  >................<
000090 90 91 92 93 94 95 96 97 98 99 9a 9b 9c 9d 9e 9f  >................<
0000a0 a0 a1 a2 a3 a4 a5 a6 a7 a8 a9 aa ab ac ad ae af  >................<
0000b0 b0 b1 b2 b3 b4 b5 b6 b7 b8 b9 ba bb bc bd be bf  >................<
0000c0 c0 c1 c2 c3 c4 c5 c6 c7 c8 c9 ca cb cc cd ce cf  >................<
0000d0 d0 d1 d2 d3 d4 d5 d6 d7 d8 d9 da db dc dd de df  >................<
0000e0 e0 e1 e2 e3 e4 e5 e6 e7 e8 e9 ea eb ec ed ee ef  >................<
0000f0 f0 f1 f2 f3 f4 f5 f6 f7 f8 f9 fa fb fc fd fe ff  >................<
000100

--
If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.

 

[robherc]

Well, now I must admit to a moronic mistake.....

After a few shots at replicating the error by modifying your .c file with my functions, I found the problem, well, at least I found the *current* problem. After my "case 255:" statement, I'd forgotten to insert a "break;" before the "default" case, so default was overwriting the output of case 255. That still doesn't tell me what was wrong BEFORE I added the default case, but it does fix my program to where its output matches 100% to what I was expecting/hoping to see when I run both test.bin, and my target files it'd been failing on before, through it.

Moral of the story: Check your code, double-check your code, triple check your code, then be prepared to feel sheepish when someone you ask for help points out your dum typo! rofl.

Thanks Salem; I'd already spent several days pulling my hair out over that, and likely would've spent a couple weeks more!

I hope this helps;
robherc