refactor(parsing): simplify emojify() and the map used by it

Signed-off-by: Rongrong <[email protected]>

docs/CHANGELOG.md

@@ -5,6 +5,7 @@
 ### Enhancements
 
 - **Print Telegram user info of bot**: Print the bot's Telegram user info when the bot is started. This is to help bot managers to find the bot's username and user ID when deploying the bot.
+- **Minor refactor**: Some internal functions have been refactored to improve performance and maintainability.
 
 ### Bug fixes
 

docs/CHANGELOG.zh.md

@@ -5,6 +5,7 @@
 ### 增强
 
 - **打印 bot 的 Telegram 用户信息**: 在 bot 启动时打印 bot 的 Telegram 用户信息。这是为了帮助 bot 管理员在部署 bot 时找到 bot 的用户名和用户 ID。
+- **次要的重构**: 重构了一些内部函数以提高性能和可维护性。
 
 ### Bug 修复
 

src/parsing/utils.py

@@ -2,7 +2,6 @@
 from typing import Optional, Sequence, Union, Final, Iterable
 
 import re
-import json
 import string
 from contextlib import suppress
 from bs4.element import Tag
@@ -11,9 +10,9 @@
 from telethon.tl.types import TypeMessageEntity
 from functools import partial
 from urllib.parse import urljoin
-from os import path
 from itertools import chain
 
+from .weibo_emojify_map import EMOJIFY_MAP
 from .. import log
 from ..aio_helper import run_async
 from ..compat import parsing_utils_html_validator_minify, INT64_T_MAX
@@ -83,10 +82,6 @@
     sorted(set(SPACES + INVALID_CHARACTERS + string.punctuation + string.whitespace))
 )
 
-# load emoji dict
-with open(path.join(path.dirname(__file__), 'emojify.json'), 'r', encoding='utf-8') as emojify_json:
-    EMOJI_DICT = json.load(emojify_json)
-
 replaceInvalidCharacter = partial(re.compile(rf'[{INVALID_CHARACTERS}]').sub, ' ')  # use initially
 replaceSpecialSpace = partial(re.compile(rf'[{SPACES[1:]}]').sub, ' ')  # use carefully
 stripBr = partial(re.compile(r'\s*<br\s*/?\s*>\s*').sub, '<br>')
@@ -121,9 +116,9 @@ def resolve_relative_link(base: Optional[str], url: Optional[str]) -> str:
 
 def emojify(xml):
     xml = emojize(xml, language='alias', variant='emoji_type')
-    for emoticon, emoji in EMOJI_DICT.items():
+    for emoticon_phrase, emoji in EMOJIFY_MAP.items():
         # emojify weibo emoticons, get all here: https://api.weibo.com/2/emotions.json?source=1362404091
-        xml = xml.replace(f'[{emoticon}]', emoji)
+        xml = xml.replace(emoticon_phrase, emoji)
     return xml
 
 

src/parsing/weibo_emojify_map.py

@@ -0,0 +1,107 @@
+from __future__ import annotations
+
+from typing import Final
+
+EMOJIFY_MAP: Final[dict] = {
+    "[微笑]": "🙂",
+    "[可爱]": "😊",
+    "[太开心]": "😆",
+    "[鼓掌]": "👏",
+    "[嘻嘻]": "😁",
+    "[哈哈]": "😄",
+    "[笑cry]": "😂",
+    "[挤眼]": "😜",
+    "[馋嘴]": "😋",
+    "[黑线]": "😑",
+    "[汗]": "😓",
+    "[哼]": "😠",
+    "[怒]": "😡",
+    "[可怜]": "🥺",
+    "[失望]": "😞",
+    "[悲伤]": "😢",
+    "[泪]": "😭",
+    "[害羞]": "😳",
+    "[爱你]": "🥰",
+    "[亲亲]": "😚",
+    "[色]": "😍",
+    "[阴险]": "😏",
+    "[偷笑]": "🤭",
+    "[酷]": "😎",
+    "[并不简单]": "🧐",
+    "[思考]": "🤔",
+    "[晕]": "😵",
+    "[骷髅]": "💀",
+    "[嘘]": "🤫",
+    "[闭嘴]": "🤐",
+    "[傻眼]": "😮",
+    "[吃惊]": "😲",
+    "[吐]": "🤮",
+    "[感冒]": "😷",
+    "[生病]": "🤒",
+    "[拜拜]": "👋",
+    "[鄙视]": "🖕",
+    "[白眼]": "🙄",
+    "[抓狂]": "😖",
+    "[怒骂]": "🤬",
+    "[钱]": "🤑",
+    "[哈欠]": "🥱",
+    "[困]": "😴",
+    "[睡]": "😪",
+    "[吃瓜]": "🍉",
+    "[酸]": "🍋",
+    "[喵喵]": "🐱",
+    "[抱抱]": "🤗",
+    "[摊手]": "🤷",
+    "[跪了]": "🧎",
+    "[鲜花]": "🌹",
+    "[给你小心心]": "💝",
+    "[心]": "❤",
+    "[伤心]": "💔",
+    "[握手]": "🤝",
+    "[赞]": "👍",
+    "[good]": "👍",
+    "[弱]": "👎",
+    "[NO]": "✋",
+    "[耶]": "✌",
+    "[拳头]": "✊",
+    "[ok]": "👌",
+    "[加油]": "💪",
+    "[haha]": "🤟",
+    "[熊猫]": "🐼",
+    "[兔子]": "🐰",
+    "[猪头]": "🐷",
+    "[太阳]": "🌞",
+    "[月亮]": "🌙",
+    "[浮云]": "☁",
+    "[下雨]": "🌧",
+    "[微风]": "🍃",
+    "[围观]": "👨‍👧‍👦",
+    "[飞机]": "✈",
+    "[照相机]": "📷",
+    "[话筒]": "🎙",
+    "[蜡烛]": "🕯",
+    "[音乐]": "🎵",
+    "[可乐]": "🥤",
+    "[干杯]": "🍻",
+    "[蛋糕]": "🎂",
+    "[礼物]": "🎁",
+    "[钟]": "⏰",
+    "[肥皂]": "🧼",
+    "[绿丝带]": "🎗",
+    "[围脖]": "🧣",
+    "[圣诞老人]": "🎅",
+    "[文明遛狗]": "🐕",
+    "[最右]": " →_→ ",
+    "[五仁月饼]": "🥮",
+    "[弗莱见钱眼开]": "🤑",
+    "[棒棒糖]": "🍭",
+    "[炸鸡腿]": "🍗",
+    "[点亮平安灯]": "🏮",
+    "[点亮橙色]": "🖐",
+    "[看涨]": "📈",
+    "[看跌]": "📉",
+    "[星星]": "⭐",
+    "[空星]": "★",
+    "[全家福]": "👪",
+    "[圆月]": "🌕"
+}