How do I print a UChar* var as a string from inside of gdb?

Question

I've been doing some work on an unfamiliar codebase that uses UChar* as strings. Uchars are defined as follows (at least according to gdb)

(gdb) ptype UChar
type = short unsigned int

However, when I try to print these in gdb, I just get the address. I can also index into the pointer and retrieve the values of each character.

Is there any way to print a variable of type UChar* from within gdb and get back a meaningful string?

Also, this is on OS X, if that makes any difference.

Answer 1

You first need to figure out what a UChar actually represents. It is likely UTF-16 or UCS-2 (but BE or LE?). Once you determine this, you want to provide (you can probably use existing code, such as iconv) a debug method to convert to UTF-8. See http://www.skynet.ie/~caolan/TechTexts/GdbUnicodePrinting.html for details.

Answer 2

if it is an ascii string, you might try to tell gdb to reinterpret:

(gdb) print (char*) theUcharPtr

Answer 3

print is the same as x; x/1s 0x1234 -- will print out that location in memory as a string, if you keep hitting carrage return, it will print the next line... etc...

If you want to monitor something continually, use display/ with the same format specifier as x (print). "display/1s 0x1234" then every time you break via a breakpoint or a single step, you will see the information you configured print out .. updated etc...

Answer 4

Just define this command in your .gdbinit and type uc varname (uc will likely work as a short form for the ucharprint command you define)

define ucharprint
echo "
set $c = (unsigned short*)$arg0
while ( *$c )
  if ( *$c > 0x7f )
    printf "[%x]", *$c
  else
    printf "%c", *$c
  end
  set $c++
end
echo "\n
end

You don't need to worry about endianness since each unsigned short in your UTF-16 UChar type holds a code point (or half surrogate) as a native binary integer.