build_emoji: Generate codepoint_to_name.json file.

Store the `codepoint_to_name` map in a JSON file which can be used
by the rest of codebase(zerver) to convert emoji codepoints to
canonical names.

tools/setup/emoji/build_emoji

@@ -16,8 +16,8 @@ from typing import Dict, Text, Union
 from os.path import dirname
 from PIL import Image, ImageDraw, ImageFont
 
-from emoji_setup_utils import generate_emoji_catalog, emoji_names_for_picker, \
+from emoji_setup_utils import generate_emoji_catalog, generate_codepoint_to_name_map, \
-    EMOJISETS
+    emoji_names_for_picker, EMOJISETS
 
 ZULIP_PATH = os.path.join(os.path.dirname(os.path.abspath(__file__)), '../../../')
 sys.path.append(ZULIP_PATH)
@@ -205,6 +205,10 @@ def dump_emojis(cache_path):
     emoji_data = ujson.load(open(os.path.join(EMOJI_DATA_PATH, 'emoji.json')))
     emoji_catalog = generate_emoji_catalog(emoji_data)
 
+    UNIFIED_REACTIONS_PATH = os.path.join(ZULIP_PATH, 'zerver', 'management', 'data', 'unified_reactions.json')
+    with open(UNIFIED_REACTIONS_PATH) as unified_reactions_file:
+        unified_reactions_data = ujson.load(unified_reactions_file)
+
     os.chdir(EMOJI_DUMP_DIR_PATH)
 
     try:
@@ -319,5 +323,12 @@ def dump_emojis(cache_path):
     name_to_codepoint_file.write(ujson.dumps(emoji_map))
     name_to_codepoint_file.close()
 
+    CODEPOINT_TO_NAME_PATH = os.path.join(cache_path, 'codepoint_to_name.json')
+    codepoint_to_name_file = open(CODEPOINT_TO_NAME_PATH, 'w')
+    codepoint_to_name = generate_codepoint_to_name_map(names, unified_reactions_data)
+
+    codepoint_to_name_file.write(ujson.dumps(codepoint_to_name))
+    codepoint_to_name_file.close()
+
 if __name__ == "__main__":
     main()

tools/setup/emoji/emoji_setup_utils.py

@@ -256,3 +256,14 @@ def emoji_is_universal(emoji_dict):
         if not emoji_dict['has_img_' + emoji_set]:
             return False
     return True
+
+def generate_codepoint_to_name_map(names, unified_reactions_data):
+    # type: (List[str], Dict[Text, Text]) -> Dict[str, str]
+    # TODO: Decide canonical names. For now, using the names
+    # generated for emoji picker. In case of multiple names
+    # for the same emoji, lexicographically greater name is
+    # used, for example, `thumbs_up` is used and not `+1`.
+    codepoint_to_name = {}  # type: Dict[str, str]
+    for name in names:
+        codepoint_to_name[str(unified_reactions_data[name])] = str(name)
+    return codepoint_to_name