Python regex: Bad character range

Question

I have the next regular expression to find emojis on a text:

re.compile(u'([\U00002600-\U000027BF])|([\U0001F300-\U0001F64F])|([\U0001F680-\U0001F6FF])')

It is working well in Python 3 but in Python 2.7 I get this:

sre_constants.error: bad character range

How can I fix it to support both, Python 2.7 and Python 3?

Ricardo Branco · Accepted Answer · 2017-06-01 18:44:52Z

1

Use r'(... instead of u'(... like this:

re.compile(r'([\U00002600-\U000027BF\U0001F300-\U0001F64F\U0001F680-\U0001F6FF])')

Also note that you can specify multiple ranges inside [...]

answered Jun 1, 2017 at 18:24

Ricardo Branco

6,1091 gold badge26 silver badges32 bronze badges

1

It's interesting. I tried your solution in interactive mode and it works but when I execute the program or when I run pytest, it fails. What could be the cause?
– David Moreno García
Commented Jun 1, 2017 at 18:37
Which version of Python failed?
– Ricardo Branco
Commented Jun 1, 2017 at 18:41
The interpreter is Python 2.7.0 and that code inside a file run by python <file> works ok. It's only in my system when running the cli or running pytest locally. This is the code github.com/davidmogar/cucco/tree/executable and this is the build in Travis travis-ci.org/davidmogar/cucco/builds/238482411 Do you think that is only a problem with my system or this could happen to others? I'm running it on a Macbook.
– David Moreno García
Commented Jun 1, 2017 at 19:52
Just tried again in a virtual environment with 3.6.0 and still fails in local. The build for 3.6.0 in Travis works.
– David Moreno García
Commented Jun 1, 2017 at 19:56

Add a comment |

1 Answer 1