implement b"..." byte array literals about julia HOT 5 CLOSED

What should the escaping and interpolation behavior be?

from julia.

Escaping should be:

1. All traditional C escapes insert the byte they normally represent (in particular, \xff inserts the byte 255).
2. All Unicode escapes insert their corresponding UTF-8 byte sequence

For interpolation, we should maybe have a bytes(x) function that tries to convert objects to byte arrays and then splices the returned byte array into the middle of the byte array.

Sound reasonable?

from julia.

And I know just what to do for the default implementation of bytes:

function bytes(x)
  buf = memio()
  write(buf, x)
  takebuf_array(buf)
end

Though it would be much more efficient for byte array interpolation to build its result in a single memio by calling write on each element. In fact we could say the only difference between byte array and string interpolation is that byte arrays use write and strings use print.

from julia.

Yeah, that seems very reasonable. No reason we shouldn't do it like that.

On Tue, Jun 28, 2011 at 3:53 PM, JeffBezanson <
[email protected]>wrote:

And I know just what to do for the default implementation of bytes:

function bytes(x)
buf = memio()
write(buf, x)
takebuf_array(buf)
end

Though it would be much more efficient for byte array interpolation to
build its result in a single memio by calling write on each element. In
fact we could say the only difference between byte array and string
interpolation is that byte arrays use write and strings use print.

Reply to this email directly or view it on GitHub:
#11 (comment)

from julia.

Closed by 6c0f464.

from julia.