@Dennis oh god your avatar

@ASCII-only sys.stdin.read() will read all bytes from STDIN.

(well, on Python 3, it would be characters, not bytes, since it depends on your locale)

Yeah, it will work fine

$ echo -e \\x00 | python3 -c 'import sys;print(sys.stdin.read().encode())'
b'\x00\n'

You can't really type them

@ASCII-only Nope. When two processes are piped together, they run at the same time. The output of the first is piped to the input of the second.

(nope to the first question, that is)

If you just use echo or cat (on a relatively small input), it will probably terminate before Python even gets to the reading input step

Sure

ast.literal_eval is your friend

@NathanMerrill I agree. Dennis for mod! (oh, wait...)

@ASCII-only ast.literal_eval takes a string and outputs a Python object. It's basically the opposite of repr.

@quartata Halloween is just around the corner

@ASCII-only Don't use null bytes?

@Dennis Frightening. Children are running away screaming. Actually that might just be because of me.

@Dennis Oh please no. sys.stdin.read() exists for a reason. Need bytes instead of Unicode text? .encode(). Also, leaving dangling file references and assuming they'll get cleaned up is bad - that behavior is not consistent across implementations (PyPy doesn't automatically close file handles at the end of execution, for example).

Maybe I should make a Halloween challenge

Something fun and not terribly difficult, like (n)spooky(n+2)me

There are hundreds of resources for learning how to write good Python code :P

No, like that other site that we don't talk about

@ASCII-only Nope, never

@ASCII-only Story of my life

@ASCII-only He definitely never said that

It can. I showed you exactly how to make it work.

actually.tryitonline.net/#code=Tw&input=J1x4MDAn

@NathanMerrill Yes. One of those.

@feersum The absolute value of the difference of the sizes is a number in Z_0+. The probability that abs(new-old) == 0 is 0.

@quartata Share the code?