Error installing polyglot in python 3.5.2

Question

I want to do sentiment analysis on urdu sentences. I searched a python package Polyglot having URDU POS tagger in it. But on installing, it prompts error;

Any way out?

Answer 1

Even I struggled with this error for a long time.

Follow these steps to install polyglot on Windows:

1. download package from [https://pypi.python.org/pypi/polyglot][1] . extract into folder
2. download and install these whl files : pycld2-0.31-cp36-cp36m-win_amd64.whl, PyICU-1.9.8-cp36-cp36m-win_amd64.whl. (Also numpy - I didn't need this since I already had it installed)
3. Open setup.py in notepad and make the following changes: (need to install codecs for this)

The README.rst file throws codecs error. Replace the 2 lines with the following:

import codecs
with codecs.open('README.rst', 'r', encoding='utf-8') as readme_file:
    readme = readme_file.read()

4. install using 'python setup.py install'

I just made sure that the 'utf-8' UnicodeDecodeError is handled while reading file.

Hope this helps!!!

Answer 2

You need to install PyICU before running pip install polyglot.

You can follow PyICU Installation for Windows from https://pypi.org/project/PyICU/