Home/ Questions/Q 7183929
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-05-28T18:05:47+00:00

Here is some code that takes an int error code (scode) and tries to

  • 0

Here is some code that takes an int error code (scode) and tries to see if it fits a certain pattern. Should I be using the modulo division operator to be doing this?

const int MASK_SYNTAX_ERR = -2146827000;
if ((MASK_SYNTAX_ERR % scode) == MASK_SYNTAX_ERR)
    scriptError.GetSourceLineText(out sourceLine);

Background: I had to deduce the value of MASK_SYNTAX_ERR through observation. Here are the various syntax error codes that I observed:

// -Int Value (Formatted Value "0x{0:X8}")
-2146827281 (0x800A03EF)
-2146827279 (0x800A03F1)
-2146827280 (0x800A03F0)
-2146827283 (0x800A03ED)
-2146827284 (0x800A03EC)

Here are a couple of logic error codes for comparison:

-2146823281 (0x800A138F)
-2146823279 (0x800A1391)

(Trivia: The code itself is calling IActiveScriptError.GetSourceLineText, this is from a IActiveScriptSite.OnScriptError implementation.)

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    The bitwise & operator is what you should use, like this:
    const int MASK_SYNTAX_ERR = -2146827000;
    if (MASK_SYNTAX_ERR & scode != 0)
    scriptError.GetSourceLineText(out sourceLine);

    More about what’s going on.

    The binary representation of MASK_SYNTAX_ERR is

    1000 0000 0000 1010 0000 0101 0000 1000

    This comparison will return true for any number that has 1’s in the same positions. Take your first syntax error code, for example:

    MASK 1000 0000 0000 1010 0000 0101 0000 1000
CODE 1000 0000 0000 1010 0000 0011 1110 1111
   & 1000 0000 0000 1010 0000 0001 0000 1000 != 0

    So the mask works here. Now, comparing a logic error code:

    MASK 1000 0000 0000 1010 0000 0101 0000 1000
CODE 1000 0000 0000 1010 0001 0011 1000 1111
   & 1000 0000 0000 1010 0000 0001 0000 1000 != 0

    It also works here, which it shouldn’t. It seems like maybe your deduced mask may be wrong because it is too complicated. Looking at your syntax codes and logic codes together:

    SYNTAX 1000 0000 0000 1010 0000 0011 1110 1111
       1000 0000 0000 1010 0000 0011 1111 0001
       1000 0000 0000 1010 0000 0011 1111 0000
       1000 0000 0000 1010 0000 0011 1110 1101
       1000 0000 0000 1010 0000 0011 1110 1100
LOGIC  1000 0000 0000 1010 0001 0011 1000 1111
       1000 0000 0000 1010 0001 0011 1001 0001

    It looks like the 13th bit from the left is the key difference between logic errors and syntax errors. So you could do something like this:

    const int SYNTAX_MASK = 1 << 12;
if (scode & SYNTAX_MASK != 0)
{
    //It's a syntax error
    scriptError.GetSourceLine(out sourceLine);
}
else
{
    //It's a logic error
}

    You can use a similar analysis to figure out masks for other purposes. A lot of times, a mask is used to grab a single bit from an int, rather than something complicated like you had. It could be that the 0x800A part of the code means a certain kind of error, and the last part gives information about the error. You’ll have to do some experimentation yourself, but hopefully this will get you on the right track for doing bitmasks.

    • 0