Fix unparsing malformed entities, bump v1.10.10

telethon/extensions/html.py

@@ -174,10 +174,12 @@ def unparse(text: str, entities: Iterable[TypeMessageEntity], _offset: int = 0,
         # Otherwise we would end up with malformed text and fail to encode.
         # For example of bad input: "Hi \ud83d\ude1c"
         # https://en.wikipedia.org/wiki/UTF-16#U+010000_to_U+10FFFF
-        if '\ud800' <= text[relative_offset] <= '\udfff':
+        while (relative_offset < _length
+                and '\ud800' <= text[relative_offset] <= '\udfff'):
             relative_offset += 1
 
-        if '\ud800' <= text[relative_offset + length] <= '\udfff':
+        while (relative_offset + length < _length
+                and '\ud800' <= text[relative_offset + length] <= '\udfff'):
             length += 1
 
         entity_text = unparse(text=text[relative_offset:relative_offset + length],
@@ -222,7 +224,7 @@ def unparse(text: str, entities: Iterable[TypeMessageEntity], _offset: int = 0,
             skip_entity = True
         last_offset = relative_offset + (0 if skip_entity else length)
 
-    if last_offset < len(text) and '\ud800' <= text[last_offset] <= '\udfff':
+    while last_offset < _length and '\ud800' <= text[last_offset] <= '\udfff':
         last_offset += 1
 
     html.append(escape(text[last_offset:]))

telethon/extensions/markdown.py

@@ -189,7 +189,7 @@ def unparse(text, entities, delimiters=None, url_fmt=None):
         # Otherwise we would end up with malformed text and fail to encode.
         # For example of bad input: "Hi \ud83d\ude1c"
         # https://en.wikipedia.org/wiki/UTF-16#U+010000_to_U+10FFFF
-        if '\ud800' <= text[at] <= '\udfff':
+        while at < len(text) and '\ud800' <= text[at] <= '\udfff':
             at += 1
 
         text = text[:at] + what + text[at:]

telethon/version.py

@@ -1,3 +1,3 @@
 # Versions should comply with PEP440.
 # This line is parsed in setup.py:
-__version__ = '1.10.9'
+__version__ = '1.10.10'

tests/telethon/crypto/test_rsa.py

@@ -1,13 +1,14 @@
 """
-tests for telethon.crypto.rsa
+Tests for `telethon.crypto.rsa`.
 """
 import pytest
 
 from telethon.crypto import rsa
 
+
 @pytest.fixture
 def server_key_fp():
-    """factory to return a key, old if so chosen"""
+    """Factory to return a key, old if so chosen."""
     def _server_key_fp(old: bool):
         for fp, data in rsa._server_keys.items():
             _, old_key = data
@@ -16,22 +17,26 @@ def server_key_fp():
 
     return _server_key_fp
 
+
 def test_encryption_inv_key():
-    """test for #1324"""
+    """Test for #1324."""
     assert rsa.encrypt("invalid", b"testdata") is None
 
+
 def test_encryption_old_key(server_key_fp):
-    """test for #1324"""
+    """Test for #1324."""
     assert rsa.encrypt(server_key_fp(old=True), b"testdata") is None
 
+
 def test_encryption_allowed_old_key(server_key_fp):
     data = rsa.encrypt(server_key_fp(old=True), b"testdata", use_old=True)
-    # we can't verify the data is actually valid because we don't have
+    # We can't verify the data is actually valid because we don't have
     # the decryption keys
     assert data is not None and len(data) == 256
 
+
 def test_encryption_current_key(server_key_fp):
     data = rsa.encrypt(server_key_fp(old=False), b"testdata")
-    # we can't verify the data is actually valid because we don't have
+    # We can't verify the data is actually valid because we don't have
     # the decryption keys
     assert data is not None and len(data) == 256

tests/telethon/extensions/test_html.py

@@ -0,0 +1,38 @@
+"""
+Tests for `telethon.extensions.html`.
+"""
+from telethon.extensions import html
+from telethon.tl.types import MessageEntityBold, MessageEntityTextUrl
+
+
+def test_entity_edges():
+    """
+    Test that entities at the edges (start and end) don't crash.
+    """
+    text = 'Hello, world'
+    entities = [MessageEntityBold(0, 5), MessageEntityBold(7, 5)]
+    result = html.unparse(text, entities)
+    assert result == '<strong>Hello</strong>, <strong>world</strong>'
+
+
+def test_malformed_entities():
+    """
+    Test that malformed entity offsets from bad clients
+    don't crash and produce the expected results.
+    """
+    text = '🏆Telegram Official Android Challenge is over🏆.'
+    entities = [MessageEntityTextUrl(offset=2, length=43, url='https://example.com')]
+    result = html.unparse(text, entities)
+    assert result == '🏆<a href="https://example.com">Telegram Official Android Challenge is over🏆</a>.'
+
+
+def test_trailing_malformed_entities():
+    """
+    Similar to `test_malformed_entities`, but for the edge
+    case where the malformed entity offset is right at the end
+    (note the lack of a trailing dot in the text string).
+    """
+    text = '🏆Telegram Official Android Challenge is over🏆'
+    entities = [MessageEntityTextUrl(offset=2, length=43, url='https://example.com')]
+    result = html.unparse(text, entities)
+    assert result == '🏆<a href="https://example.com">Telegram Official Android Challenge is over🏆</a>'

tests/telethon/extensions/test_markdown.py

@@ -0,0 +1,38 @@
+"""
+Tests for `telethon.extensions.markdown`.
+"""
+from telethon.extensions import markdown
+from telethon.tl.types import MessageEntityBold, MessageEntityTextUrl
+
+
+def test_entity_edges():
+    """
+    Test that entities at the edges (start and end) don't crash.
+    """
+    text = 'Hello, world'
+    entities = [MessageEntityBold(0, 5), MessageEntityBold(7, 5)]
+    result = markdown.unparse(text, entities)
+    assert result == '**Hello**, **world**'
+
+
+def test_malformed_entities():
+    """
+    Test that malformed entity offsets from bad clients
+    don't crash and produce the expected results.
+    """
+    text = '🏆Telegram Official Android Challenge is over🏆.'
+    entities = [MessageEntityTextUrl(offset=2, length=43, url='https://example.com')]
+    result = markdown.unparse(text, entities)
+    assert result == "🏆[Telegram Official Android Challenge is over🏆](https://example.com)."
+
+
+def test_trailing_malformed_entities():
+    """
+    Similar to `test_malformed_entities`, but for the edge
+    case where the malformed entity offset is right at the end
+    (note the lack of a trailing dot in the text string).
+    """
+    text = '🏆Telegram Official Android Challenge is over🏆'
+    entities = [MessageEntityTextUrl(offset=2, length=43, url='https://example.com')]
+    result = markdown.unparse(text, entities)
+    assert result == "🏆[Telegram Official Android Challenge is over🏆](https://example.com)"