From 75d99fbb53c1bb7d46f9b9909c056b6e00eab4d8 Mon Sep 17 00:00:00 2001
From: Lonami Exo <totufals@hotmail.com>
Date: Thu, 15 Feb 2018 11:52:46 +0100
Subject: [PATCH] Fix HTML entity parsing failing when needing surrogates

---
 telethon/extensions/html.py     | 23 +++++++++++++++++++----
 telethon/extensions/markdown.py |  3 +++
 2 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/telethon/extensions/html.py b/telethon/extensions/html.py
index 8cd170cb..bcbd13cc 100644
--- a/telethon/extensions/html.py
+++ b/telethon/extensions/html.py
@@ -1,9 +1,10 @@
 """
 Simple HTML -> Telegram entity parser.
 """
+import struct
+from collections import deque
 from html import escape, unescape
 from html.parser import HTMLParser
-from collections import deque
 
 from ..tl.types import (
     MessageEntityBold, MessageEntityItalic, MessageEntityCode,
@@ -12,6 +13,18 @@ from ..tl.types import (
 )
 
 
+# Helpers from markdown.py
+def _add_surrogate(text):
+    return ''.join(
+        ''.join(chr(y) for y in struct.unpack('<HH', x.encode('utf-16le')))
+        if (0x10000 <= ord(x) <= 0x10FFFF) else x for x in text
+    )
+
+
+def _del_surrogate(text):
+    return text.encode('utf-16', 'surrogatepass').decode('utf-16')
+
+
 class HTMLToTelegramParser(HTMLParser):
     def __init__(self):
         super().__init__()
@@ -109,8 +122,8 @@ def parse(html):
     :return: a tuple consisting of (clean message, [message entities]).
     """
     parser = HTMLToTelegramParser()
-    parser.feed(html)
-    return parser.text, parser.entities
+    parser.feed(_add_surrogate(html))
+    return _del_surrogate(parser.text), parser.entities
 
 
 def unparse(text, entities):
@@ -124,6 +137,8 @@ def unparse(text, entities):
     """
     if not entities:
         return text
+
+    text = _add_surrogate(text)
     html = []
     last_offset = 0
     for entity in entities:
@@ -164,4 +179,4 @@ def unparse(text, entities):
             skip_entity = True
         last_offset = entity.offset + (0 if skip_entity else entity.length)
     html.append(text[last_offset:])
-    return ''.join(html)
+    return _del_surrogate(''.join(html))
diff --git a/telethon/extensions/markdown.py b/telethon/extensions/markdown.py
index 10327c46..92589a17 100644
--- a/telethon/extensions/markdown.py
+++ b/telethon/extensions/markdown.py
@@ -152,6 +152,9 @@ def unparse(text, entities, delimiters=None, url_fmt=None):
     :param entities: the MessageEntity's applied to the text.
     :return: a markdown-like text representing the combination of both inputs.
     """
+    if not entities:
+        return text
+
     if not delimiters:
         if delimiters is not None:
             return text