parsing an hex value with strconv

[Rodrigo Moraes]

Hey. With weekly.2012-02-22:

i, _ := strconv.ParseInt("FF", 16, 8) // 127
i, _ := strconv.ParseInt("FF", 16, 16) // 255

I'd expect that both would be 255. Why this happened?

-- rodrigo

[minux]

Because ParseInt parses *signed* numbers, and 8-bit signed integer could only represent

integers in range [-128, 127].

I think you really want strconv.ParseUint.

[Rodrigo Moraes]

On Feb 27, 1:47 pm, minux wrote:
> Because ParseInt parses *signed* numbers, and 8-bit signed integer could
> only represent
> integers in range [-128, 127].
>
> I think you really want strconv.ParseUint.

Ah, doh, of course! Thank you. :)

-- rodrigo

[Dustin]

According to the docs:

"The bitSize argument specifies the integer type that the result must fit into. Bit sizes 0, 8, 16, 32, and 64 correspond to int, int8, int16, int32, and int64."

255 isn't a valid 8 bit signed integer.  I would expect -1.

[minux]

It's out of range, so ErrRange is returned, and the document doesn't say anything about

the value returned when error occurred.

[Dustin]

On Monday, February 27, 2012 9:01:47 AM UTC-8, minux wrote:

255 isn't a valid 8 bit signed integer.  I would expect -1.

It's out of range, so ErrRange is returned, and the document doesn't say anything about

the value returned when error occurred.

That's an excellent point (always check the error).  I was looking at 0xFF as eight one bits, which is -1.  I guess I could argue with the parser as to which makes more sense, but I'll let it win.