如何使用Python从unicode字符串中切片子字符串？

Question

I have a unicode string as a result : u'splunk>\\xae\'

How can I get the substring 'uf001'

as a simple string in python?

Answer 1

The characters uf001 are not actually present in the string, so you can't just slice them off. You can do

repr(s)[-6:-1]

or

'u' + hex(ord(s[-1]))[2:]

Answer 2

Since you want the actual string (as seen from comments) , just get the last character [-1] index , Example -

>>> a = u'splunk>\xae\uf001'
>>> print(a)
splunk>®ï€
>>> a[-1]
'\uf001'
>>> print(a[-1])
ï€

If you want the unicode representation ( \ ) , then take repr(a[-1]) , Example -

>>> repr(a[-1])
"'\\uf001'"

\ is a single unicode character (not multiple strings) , so you can directly get that character as above.

You see \ because you are checking the results of repr() on the string, if you print it, or use it somewhere else (like for files, etc) it will be the correct \ character.

Answer 3

u'' it is how a Unicode string is represented in Python source code. REPL uses this representation by default to display unicode objects:

>>> u'splunk>\xae\uf001'
u'splunk>\xae\uf001'
>>> print(u'splunk>\xae\uf001')
splunk>®
>>> print(u'splunk>\xae\uf001'[-1])


If your terminal is not configured to display Unicode or if you are on a narrow build (eg, it is likely for Python 2 on Windows) then the result may be different.

Unicode string is an immutable sequence of Unicode codepoints in Python. len(u'\') == 1 : it does not contain uf001 (5 characters) in it. You could write it as u'' (it is necessary to declare the character encoding of your source file on Python 2 if you use non-ascii characters):

>>> u'\uf001' == u''
True

It is just a different way to represent exactly the same Unicode character (a single codepoint in this case).

Note: some user-perceived characters may span several Unicode codepoints eg:

>>> import unicodedata
>>> unicodedata.normalize('NFKD', u'ё')
u'\u0435\u0308'
>>> print(unicodedata.normalize('NFKD', u'ё'))
ё