Realtime code analysis stops working when non-ASCII characters are used in strings

Hi everyone. Searched for the problem I’m about to report, and didn’t found anything. Please excuse me about any possible mistakes, as I’m not used to Github issue reporting.

Problem Description

Code highlighting and analysis suddenly stops as soon as you type any latin1 character inside the editor. Removing the character resumes the normal process.

(PLEASE NOTICE THE DIFFERENCE IN THE HIGHLIGHTED “ê”) As a Brazilian user (many ê, â, ã etc needs), this is somewhat frustrating.

Package Versions

• Spyder: 3.2.8
• Python: 3.6
• PyQt: 5.9.2
• Operating System: W10

Dependencies

IPython >=4.0    :  6.3.0 (OK)
cython >=0.21    :  0.28.2 (OK)
jedi >=0.9.0     :  0.11.1 (OK)
nbconvert >=4.0  :  5.3.1 (OK)
numpy >=1.7      :  1.14.2 (OK)
pandas >=0.13.1  :  None (NOK)
pycodestyle >=2.3:  2.3.1 (OK)
pyflakes >=0.6.0 :  1.6.0 (OK)
pygments >=2.0   :  2.2.0 (OK)
pylint >=0.25    :  1.8.3 (OK)
qtconsole >=4.2.0:  4.3.1 (OK)
rope >=0.9.4     :  0.10.7 (OK)
sphinx >=0.6.6   :  1.7.2 (OK)
sympy >=0.7.3    :  None (NOK)

[ccordoba12]

Yes, please.

[dalthviz]

@ccordoba12 we are already using enconding.get_coding here, should I change the behavior of the function to make possible the use of the chardet fallback directly and handle the expection here?