iamaziz / ar_wordcloud

A tiny wrapper for Arabic WordCloud plots

^{✨ May, 17 2020 ✨}

ar_wordcloud

A tiny wrapper around wordcloud.WordCloud to support WordCloud with Arabic text.

Installation:

• $ pip install ar_wordcloud

Usage:

Also, see the notebook example.

Import and instantiate the main class:

from ar_wordcloud import ArabicWordCloud

awc = ArabicWordCloud(background_color="white")

Then, either build a word cloud from a raw text string:

t = 'أهلاً وسهلا، اللغة العربية جميلة'
wc = awc.from_text(t)
wc.to_file('hello.png')

Or, build a word cloud from a dictionary:

d = {"مرحبا": 12, "برمج": 8, "من": 10, "رمضان كريم": 15, "العيد قرب": 12, "اهلين": 1, "كورونا": 3}
dict_wc = awc.from_dict(d, ignore_stopwords=True)
awc.plot(dict_wc, title="كل عام وانتم بخير", width=2, height=2)

Or, read text from a file:

wc = awc.from_file('examples/arabic.txt')
wc.to_file('examples/arabic.png')

---

Using a mask

Also, you can use a mask:

from ar_wordcloud import ArabicWordCloud
from ar_wordcloud.utils import read_mask_image

mask_img = read_mask_image() # NOTE: pass `mask_img_url` param to use a different mask, the default is a heart 
awc = ArabicWordCloud(background_color="white", mask=mask_img, contour_width=4, scale=0.5)

awc.from_file('examples/arabic.txt').to_image()

---

Multiple fonts support

It's also possible to choose a different Arabic font:

from ar_wordcloud import ArabicWordCloud

awc = ArabicWordCloud(font='NotoSansArabic-ExtraBold.ttf')

t = f"عيدفطر2020 سعيد، كل عام وانتم بخير"
awc.from_text(t).to_image()

# To see the available fonts:
print(awc.fonts.available_fonts)

If the entered font name is not correct, there'll be a friendly assertion message:

AssertionError: بالله أتأكد ان اسم الخط المُدخل صحيح
Please make sure the selected font name is correct!

---

TODO:

• support mask plots
• clean input text (remove punct .. etc)
• support input from file e.g. awc.from_file(fname)
• browse available fonts and select favorite
• fix English font
• cli support

PRs are welcome, thanks 🙏

---

Credit: this repo was born from the discussion in this PR.