Add support for Japanese

Author: gachiemchiep

MANIFEST.in

@@ -2,6 +2,7 @@ include README.md
 include LICENSE
 include trdg/fonts/latin/*
 include trdg/fonts/cn/*
+include trdg/fonts/ja/*
 include trdg/images/*
 include trdg/dicts/*
 include trdg/texts/*

README.md

@@ -164,6 +164,15 @@ Simplified:
 
 ![28](samples/28.jpg "1")
 
+## Create images with Japanese text 
+
+It is simple! Just do `trdg -l ja -c 1000 -w 5`!
+
+Output 
+
+![29](samples/29.jpg "2")
+
+
 ## Add new fonts
 
 The script picks a font at random from the *fonts* directory.
@@ -173,6 +182,7 @@ The script picks a font at random from the *fonts* directory.
 | fonts/latin | English, French, Spanish, German |
 | fonts/cn | Chinese |
 | fonts/ko | Korean |
+| fonts/ja | Japanese |
 
 Simply add/remove fonts until you get the desired output.
 

samples/29.jpg

@@ -1095,6 +1095,13 @@ def test_generate_chinese_string(self):
 
         self.assertTrue(all([l in cn_chars for l in s]))
 
+    def test_generate_japanese_string(self):
+        s = create_strings_randomly(1, False, 1, True, False, False, "ja")[0]
+
+        ja_chars = [chr(i) for i in range(12288, 12543)] + [chr(i) for i in range(65280, 65519)] + [chr(i) for i in range(19968, 40908)]
+
+        self.assertTrue(all([l in ja_chars for l in s]))
+
     def test_generate_data_with_white_background(self):
         background_generator.plain_white(64, 128).convert("RGB").save(
             "tests/out/white_background.jpg"

tests.py

@@ -1,7 +1,7 @@
 import os
 import random as rnd
 
-from PIL import Image, ImageFilter
+from PIL import Image, ImageFilter, ImageStat
 
 from trdg import computer_text_generator, background_generator, distorsion_generator
 
@@ -174,6 +174,26 @@ def generate(
             "RGB", (background_width, background_height), (0, 0, 0)
         )
 
+        ##############################################################
+        # Comparing average pixel value of text and background image #
+        ##############################################################
+        try:
+            resized_img_st = ImageStat.Stat(resized_img, resized_mask.split()[2])
+            background_img_st = ImageStat.Stat(background_img) 
+
+            resized_img_px_mean = sum(resized_img_st.mean[:2]) / 3
+            background_img_px_mean = sum(background_img_st.mean) / 3
+
+            if abs(resized_img_px_mean - background_img_px_mean) < 15:
+                print("value of mean pixel is too similar. Ignore this image")
+
+                print("resized_img_st \n {}".format(resized_img_st.mean))
+                print("background_img_st \n {}".format(background_img_st.mean))
+
+                return
+        except Exception as err:
+            return
+
         #############################
         # Place text with alignment #
         #############################

trdg/data_generator.py

@@ -51,7 +51,7 @@ def parse_arguments():
         "--language",
         type=str,
         nargs="?",
-        help="The language to use, should be fr (French), en (English), es (Spanish), de (German), ar (Arabic), cn (Chinese), or hi (Hindi)",
+        help="The language to use, should be fr (French), en (English), es (Spanish), de (German), ar (Arabic), cn (Chinese), ja (Japanese) or hi (Hindi)",
         default="en",
     )
     parser.add_argument(

trdg/fonts/ja/TakaoMincho.ttf

@@ -96,6 +96,23 @@ def create_strings_randomly(length, allow_variable, count, let, num, sym, lang):
             pool += "".join(
                 [chr(i) for i in range(19968, 40908)]
             )  # Unicode range of CHK characters
+        elif lang == "ja":
+            pool += "".join(
+                [chr(i) for i in range(12288, 12351)]
+            )   # unicode range for japanese-style punctuation
+            pool += "".join(
+                [chr(i) for i in range(12352, 12447)]
+            )   # unicode range for Hiragana
+            pool += "".join(
+                [chr(i) for i in range(12448, 12543)]
+            )   # unicode range for Katakana
+            pool += "".join(
+                [chr(i) for i in range(65280, 65519)]
+            )   # unicode range for Full-width roman characters and half-width katakana 
+            pool += "".join(
+                [chr(i) for i in range(19968, 40908)]
+            )   # unicode range for common and uncommon kanji
+            # https://stackoverflow.com/questions/19899554/unicode-range-for-japanese
         else:
             pool += string.ascii_letters
     if num:
@@ -106,6 +123,9 @@ def create_strings_randomly(length, allow_variable, count, let, num, sym, lang):
     if lang == "cn":
         min_seq_len = 1
         max_seq_len = 2
+    elif lang == "ja":
+        min_seq_len = 1
+        max_seq_len = 2
     else:
         min_seq_len = 2
         max_seq_len = 10