CODEWITHSUNDEEP
pythonhtml
Aamir Rind
Aamir Rind

Reputation: 39709

ValueError: unichr() arg not in range(0x10000) (narrow Python build)

I am trying to convert the html entity to unichar, the html entity is 󮠖 when i try to do the following:

unichr(int(976918))

I got error that:

ValueError: unichr() arg not in range(0x10000) (narrow Python build)

seems like it is out of range conversion for unichar.

Upvotes: 19

Views: 10903

Answers (3)

Mark Ransom
Mark Ransom

Reputation: 308530

Here's an alternate workaround that I developed with the struct module.

def unichar(i):
    try:
        return unichr(i)
    except ValueError:
        return struct.pack('i', i).decode('utf-32')

>>> unichar(int('976918'))
u'\U000ee816'

Upvotes: 13

Eryk Sun
Eryk Sun

Reputation: 34300

You can decode a string that has a Unicode escape (\U followed by 8 hex digits, zero-padded) using the "unicode-escape" encoding: 

>>> s = "\\U%08x" % 976918
>>> s
'\\U000ee816'

>>> c = s.decode('unicode-escape')
>>> c
u'\U000ee816'

On a narrow build it's stored as a UTF-16 surrogate pair:

>>> list(c)
[u'\udb7a', u'\udc16']

This surrogate pair is processed correctly as a code unit during encoding:

>>> c.encode('utf-8')
'\xf3\xae\xa0\x96'

>>> '\xf3\xae\xa0\x96'.decode('utf-8')
u'\U000ee816'

Upvotes: 29

agf
agf

Reputation: 176980

In order for this to work, you either need to build Python yourself, specifying 

./configure --enable-unicode=ucs4

before compiling, or else you need to move to Python 3.

Even if you do this, there are apparently problems on Windows, which will be fixed in the next version of Python (3.3).

Upvotes: 6

Related Questions