I want to use jsctypes in a js protocol component (@mozilla.org/network/protocol;1?name=). For this reasons I need to deal with binary strings which are coming from the C/C++ code as char** (or ctypes.char.ptr given by address as parameter). Unfortunately the CData readString() method is unadapted (UTF8 and null terminated assumptions) and there is no method to access quickly the underlying byte buffer. So, I have found no efficient way to access the data. It would be great to modify readString() to allow that or do build a method like data([length]).

I need something along these lines for reading Unicode from files by chunks. Emmanuel, what do you think of merging this bug with bug 552551?

Emmanuel, actually, on second thought, I am not sure of how this relates to readString. Doesn't casting your ctypes.char.ptr to an array do what you need?

I have tried to cast the char.ptr to a uint8 array and it still crashes when accessing the individual elements is there no way to get binary data? Heres my testing code var teststring = lib.declare("teststring", ctypes.default_abi, //ctypes.stdcall_abi, // call ABI ctypes.int64_t, // return type ctypes.int32_t.ptr, // buffer size ctypes.char.ptr //ctypes.char.ptr.array() // Buffer ); var bufferSize = new ctypes.int32_t(10).address(); var buffer = new ctypes.char(); var bufferPtr = buffer.address(); var ret = teststring(bufferSize, bufferPtr); var hold = ctypes.cast( bufferPtr, ctypes.ArrayType(ctypes.uint8_t, parseInt(bufferSize.contents)).ptr ); console.error("Hold array length: "+hold.contents.length); var output = ""; for(var i=0; i<bufferSize.contents; i++){ output+=""+hold.contents[i]; } console.error("Output: "+output);

I can go directly to a Uint8Array with this var byteArray = new Uint8Array(buffer); I think I was trying to hard with the ctypes casting lol