101

I'm looking for a formatted byte string literal. Specifically, something equivalent to

name = "Hello" bytes(f"Some format string {name}") 

Possibly something like fb"Some format string {name}".

Does such a thing exist?

1
  • 7
    I don't think so. bytes don't even have a .format method, so, I'd be surprised if they had f-string equivalents. The closest you'll get is bytes formatting Commented Jul 27, 2017 at 20:33

6 Answers 6

79

No. The idea is explicitly dismissed in the PEP:

For the same reason that we don't support bytes.format(), you may not combine 'f' with 'b' string literals. The primary problem is that an object's __format__() method may return Unicode data that is not compatible with a bytes string.

Binary f-strings would first require a solution for bytes.format(). This idea has been proposed in the past, most recently in PEP 461. The discussions of such a feature usually suggest either

  • adding a method such as __bformat__() so an object can control how it is converted to bytes, or

  • having bytes.format() not be as general purpose or extensible as str.format().

Both of these remain as options in the future, if such functionality is desired.

Sign up to request clarification or add additional context in comments.

Comments

46

In 3.6+ you can do:

>>> a = 123 >>> f'{a}'.encode() b'123' 

2 Comments

If you want to use bytes format, it's probably because the data you want to format (a here) is bytes. Decoding that to (potentially invalid) unicode and then going back to bytes is asking for trouble.
It's not potentially invalid, it's most likely not what you want: If awas of type bytes then f{a}.encode() would give b"b'123'"
11

You were actually super close in your suggestion; if you add an encoding kwarg to your bytes() call, then you get the desired behavior:

>>> name = "Hello" >>> bytes(f"Some format string {name}", encoding="utf-8") b'Some format string Hello' 

Caveat: This works in 3.8 for me, but note at the bottom of the Bytes Object headline in the docs seem to suggest that this should work with any method of string formatting in all of 3.x (using str.format() for versions <3.6 since that's when f-strings were added, but the OP specifically asks about 3.6+).

Comments

8

From python 3.6.2 this percent formatting for bytes works for some use cases:

print(b"Some stuff %a. Some other stuff" % my_byte_or_unicode_string)

But as AXO commented:

This is not the same. %a (or %r) will give the representation of the string, not the string iteself. For example b'%a' % b'bytes' will give b"b'bytes'", not b'bytes'.

Which may or may not matter depending on if you need to just present the formatted byte_or_unicode_string in a UI or if you potentially need to do further manipulation.

3 Comments

This is not the same. %a (or %r) will give the representation of the string, not the string iteself. For example b'%a' % b'bytes' will give b"b'bytes'", not b'bytes'.
I think you meant %s and my_byte_string, e.g. following AXO's example, b'%s' % b'bytes' -> b'bytes'
Documentation is at docs.python.org/3/library/… . Note that %b is more explicit than %s for formatting a bytes argument.
4

As noted here, you can format this way:

>>> name = b"Hello" >>> b"Some format string %b World" % name b'Some format string Hello World' 

You can see more details in PEP 461

Note that in your example you could simply do something like:

>>> name = b"Hello" >>> b"Some format string " + name b'Some format string Hello' 

Comments

-4

This was one of the bigger changes made from python 2 to python3. They handle unicode and strings differently.

This s how you'd convert to bytes.

string = "some string format" string.encode() print(string) 

This is how you'd decode to string.

string.decode() 

I had a better appreciation for the difference between Python 2 versus 3 change to unicode through this coursera lecture by Charles Severence. You can watch the entire 17 minute video or fast forward to somewhere around 10:30 if you want to get to the differences between python 2 and 3 and how they handle characters and specifically unicode.

I understand your actual question is how you could format a string that has both strings and bytes.

inBytes = b"testing" inString = 'Hello' type(inString) #This will yield <class 'str'> type(inBytes) #this will yield <class 'bytes'> 

Here you could see that I have a string a variable and a bytes variable.

This is how you would combine a byte and string into one string.

formattedString=(inString + ' ' + inBytes.encode()) 

2 Comments

I appreciate your time and efforts, but your answer is irrelevant to the question.
I understood Enrico's question specifically how he could convert bytes and strings in Python 3, an area which changed significantly in the new version. My answer was meant to help him understand how that change happened and how he could handle a similar operation in Python 3. That's why I feel this response is relevant to his question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.