Enable multiple font use within each text

examples/mix_fonts.py

@@ -0,0 +1,88 @@
+#!/usr/bin/env python
+"""
+Mix Fonts
+=========
+
+English, CJK, and emoji are mixed both across different words and
+within each word. Fonts are aligned at their baselines.
+
+"""
+import os
+import random
+
+import emoji
+import matplotlib.pyplot as plt
+import regex as re
+from fontTools.ttLib import TTFont
+
+from wordcloud import MixedFontPattern
+from wordcloud import WordCloud
+
+
+def get_supported_chars(font_path):
+    supported_chars = []
+    with TTFont(font_path, 0, allowVID=0,
+                ignoreDecompileErrors=True,
+                fontNumber=-1) as ttf:
+        for table in ttf["cmap"].tables:
+            for glyph in list(table.cmap.items()):
+                try:
+                    char = chr(glyph[0])
+                    supported_chars.append(char)
+                except Exception:
+                    continue
+    return supported_chars
+
+
+font_emoji = "fonts/Symbola/Symbola.ttf"
+font_cjk = "fonts/SourceHanSerif/SourceHanSerifK-Light.otf"
+font_paths = [
+    MixedFontPattern(re.compile(pattern), path) for pattern, path in [
+        (r"[\p{Emoji=Yes}\p{Emoji_Presentation=Yes}]+", font_emoji),
+        (r"[\p{Han}\p{Katakana}\p{Hiragana}]+", font_cjk),
+    ]
+]
+
+# construct word frequency dict.
+wordfreq = {}
+supported = get_supported_chars(font_emoji)
+emojis = list(v for v in emoji.EMOJI_UNICODE.values() if v in supported)
+wordfreq.update({random.choice(emojis): random.randint(10, 40)
+                 for _ in range(200)})
+wordfreq.update({
+    emoji.emojize(word, use_aliases=True): n for word, n in {
+        u"愛Myミィ": 75,
+        u"I:beating_heart:紐育": 75,
+        u"We:growing_heart:おんがく": 75,
+        u"男": 90,
+        u"女": 90,
+        u":man_dancing:": 90,
+        u":woman_dancing:": 90,
+        u":couple_with_heart:": 90,
+        u"愛": 90,
+        u"love": 90,
+        u"舞": 90,
+        u"dance": 90,
+        u":musical_note:": 90,
+        u":musical_notes:": 90,
+        u":musical_score:": 45,
+    }.items()
+})
+
+wc = WordCloud(width=600,
+               height=600,
+               scale=1.5,
+               relative_scaling=0.6,
+               repeat=False,
+               max_words=len(wordfreq),
+               font_path=font_paths)
+wc.generate_from_frequencies(wordfreq)
+
+# generate an svg outout.
+with open(os.path.splitext(os.path.basename(__file__))[0] + ".svg", 'w') as f:
+    f.write(wc.to_svg(embed_font=True))
+
+plt.figure(figsize=(8, 8))
+plt.imshow(wc, interpolation="bilinear")
+plt.axis("off")
+plt.show()

examples/mix_fonts_random.py

@@ -0,0 +1,40 @@
+#!/usr/bin/env python
+"""
+Mix Fonts with Random Draw Example
+==================================
+
+Generating a word cloud from the US constitution. The words are
+rendered with a font randomly picked from a list.
+
+"""
+import os
+
+import matplotlib.pyplot as plt
+
+from wordcloud import WordCloud
+
+
+# get data directory (using getcwd() is needed to support running
+# example in generated IPython notebook).
+d = os.path.dirname(__file__) if "__file__" in locals() else os.getcwd()
+
+# read the whole text.
+text = open(os.path.join(d, 'constitution.txt')).read()
+
+# provide paths to fonts for random draw.
+font_paths = [
+    "../wordcloud/DroidSansMono.ttf",
+    "fonts/Symbola/Symbola.ttf",
+    "fonts/SourceHanSerif/SourceHanSerifK-Light.otf"
+]
+
+# generate a word cloud image.
+wc = WordCloud(width=1200, height=600, font_path=font_paths).generate(text)
+
+# generate an svg outout.
+with open(os.path.splitext(os.path.basename(__file__))[0] + ".svg", 'w') as f:
+    f.write(wc.to_svg(embed_font=True))
+
+plt.imshow(wc, interpolation='bilinear')
+plt.axis("off")
+plt.show()

examples/mix_fonts_random_and_match.py

@@ -0,0 +1,60 @@
+#!/usr/bin/env python
+"""
+Mix Fonts with Random Draw Example and Match
+============================================
+
+Generating a word cloud from the US constitution. The words are
+rendered with a font randomly picked from a list except for CJK
+characters, to which a CJK font is applied.
+
+"""
+import os
+import regex as re
+from collections import defaultdict
+
+import matplotlib.pyplot as plt
+
+from wordcloud import MixedFontPattern
+from wordcloud import STOPWORDS
+from wordcloud import WordCloud
+
+
+# get data directory (using getcwd() is needed to support running
+# example in generated IPython notebook).
+d = os.path.dirname(__file__) if "__file__" in locals() else os.getcwd()
+
+# read the whole text.
+text = open(os.path.join(d, 'constitution.txt')).read()
+
+wordfreq = defaultdict(int)
+for word in text.split():
+    word = word.strip('.,')
+    if word.lower() not in STOPWORDS:
+        wordfreq[word] += 1
+
+# add some cjk language entries for demonstrating font matching
+for phrase, n in {'アメリカ': 40, '憲法': 35, '大統領': 32, '合衆国': 30,
+                  'あめりか': 30, 'けんぽう': 25, 'だいとうりょう': 23,
+                  'がっしゅうこく': 20}.items():
+    wordfreq[phrase] += n
+
+# provide paths to fonts for random draw plus cjk font when word matches
+font_paths = [
+    "../wordcloud/DroidSansMono.ttf",
+    "fonts/Symbola/Symbola.ttf",
+    "fonts/SourceHanSerif/SourceHanSerifK-Light.otf",
+    MixedFontPattern(re.compile(r"[\p{Han}\p{Katakana}\p{Hiragana}]+"),
+                     "fonts/SourceHanSerif/SourceHanSerifK-Light.otf")
+]
+
+# generate a word cloud image.
+wc = WordCloud(width=1200, height=600, font_path=font_paths)
+wc.generate_from_frequencies(wordfreq)
+
+# generate an svg outout.
+with open(os.path.splitext(os.path.basename(__file__))[0] + ".svg", 'w') as f:
+    f.write(wc.to_svg(embed_font=True))
+
+plt.imshow(wc, interpolation='bilinear')
+plt.axis("off")
+plt.show()

test/test_wordcloud.py

@@ -351,14 +351,17 @@ def test_recolor_too_small_set_default():
 
 def test_small_canvas():
     # check font size fallback works on small canvas
-    WordCloud(max_words=50, width=20, height=20).generate(THIS)
+    wc = WordCloud(max_words=50, width=21, height=21)
+    wc.generate(THIS)
+    assert len(wc.layout_) == 1
 
 
 def test_tiny_canvas():
     # check exception if canvas too small for fallback
     w = WordCloud(max_words=50, width=1, height=1)
     with pytest.raises(ValueError, match="Couldn't find space to draw"):
         w.generate(THIS)
+    assert len(w.layout_) == 0
 
 
 def test_coloring_black_works():
@@ -382,7 +385,7 @@ def test_repeat():
     # all frequencies are 1
     assert len(wc.words_) == 3
     assert_array_equal(list(wc.words_.values()), 1)
-    frequencies = [w[0][1] for w in wc.layout_]
+    frequencies = [w.frequency for w in wc.layout_]
     assert_array_equal(frequencies, 1)
     repetition_text = "Some short text with text"
     wc = WordCloud(max_words=52, stopwords=[], repeat=True)
@@ -392,7 +395,7 @@ def test_repeat():
     assert wc.words_['text'] == 1
     assert wc.words_['with'] == .5
     assert len(wc.layout_), wc.max_words
-    frequencies = [w[0][1] for w in wc.layout_]
+    frequencies = [w.frequency for w in wc.layout_]
     # check that frequencies are sorted
     assert np.all(np.diff(frequencies) <= 0)
 
@@ -403,4 +406,4 @@ def test_zero_frequencies():
 
     word_cloud.generate_from_frequencies({'test': 1, 'test1': 0, 'test2': 0})
     assert len(word_cloud.layout_) == 1
-    assert word_cloud.layout_[0][0][0] == 'test'
+    assert word_cloud.layout_[0].word == 'test'

test/test_wordcloud_cli.py

@@ -6,7 +6,10 @@
 import wordcloud as wc
 from wordcloud import wordcloud_cli as cli
 
-from mock import patch
+try:
+    from unittest.mock import patch
+except ImportError:
+    from mock import patch
 import pytest
 
 import matplotlib

wordcloud/__init__.py

@@ -1,8 +1,8 @@
-from .wordcloud import (WordCloud, STOPWORDS, random_color_func,
+from .wordcloud import (WordCloud, STOPWORDS, MixedFontPattern, random_color_func,
                         get_single_color_func)
 from .color_from_image import ImageColorGenerator
 
-__all__ = ['WordCloud', 'STOPWORDS', 'random_color_func',
+__all__ = ['WordCloud', 'STOPWORDS', 'MixedFontPattern', 'random_color_func',
            'get_single_color_func', 'ImageColorGenerator',
            '__version__']