This would presumaby support the (read-only part of the) buffer API so search would be covered. I don't see a use case for replace. Alternatively, you could always specify Latin-1 as the encoding and convert it that way -- I don't think there's any input that can cause Latin-1 decoding to fail. On 10/3/05, Antoine Pitrou <solipsis@pitrou.net> wrote:
Le lundi 03 octobre 2005 à 14:02 -0700, Guido van Rossum a écrit :
On 10/3/05, Antoine Pitrou <solipsis@pitrou.net> wrote:
Could the "bytes" type be just the same as the current "str" type but without the implicit unicode conversion ? Or am I missing some desired functionality ?
No. It will be a mutable array of bytes. It will intentionally resemble strings as little as possible. There won't be a literal for it.
Thinking about it, it may have to offer the search and replace facilities offered by strings (including regular expressions).
Here is an use case : say I'm reading an HTML file (or receiving it over the network). Since the character encoding can be specified in the HTML file itself (in the <head>...</head>), I must first receive it as a bytes object. But then I must fetch the encoding information from the HTML header: therefore I must use some string ops on the bytes object to parse this information. Only after I have discovered the encoding, can I finally convert the bytes object to a text string.
Or would there be another way to do it?
_______________________________________________ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/guido%40python.org
-- --Guido van Rossum (home page: http://www.python.org/~guido/)