WL#15212: Update collation mappings

MySQL Server 8.0.30 has added new collations and renamed some.

Support for new utf8mb4 collations has been added, and all existing
utf8_* collations have been renamed to utf8mb3_*.

This worklog updates the charsets.py module accordingly, adds the 5.7
server character set mapping, and makes the necessary adjustments to
keep backwards compatibility.

See changes introduced by WL#14307, BUG#33787300 and BUG#33635120.

Change-Id: If21839728fb52a5f82fc4c1b4f83d230f178046c

Author: nmariz

Diff for: CHANGES.txt

@@ -11,6 +11,7 @@ Full release notes:
 v8.0.30
 =======
 
+- WL#15212: Update collation mappings
 - WL#15151: Increase to 88 characters per line
 - WL#15137: Fix linting issues
 - WL#15035: Enforce PEP 7 and PEP 8 coding style

Diff for: lib/mysql/connector/abstracts.py

@@ -976,7 +976,7 @@ def python_charset(self):
         Returns a string.
         """
         encoding = CharacterSet.get_info(self._charset_id)[0]
-        if encoding in ("utf8mb4", "binary"):
+        if encoding in ("utf8mb4", "utf8mb3", "binary"):
             return "utf8"
         return encoding
 

Diff for: lib/mysql/connector/charsets.py

@@ -44,6 +44,7 @@
 from .authentication import get_auth_plugin
 from .constants import (
     NET_BUFFER_LENGTH,
+    CharacterSet,
     ClientFlag,
     FieldType,
     ServerCmd,
@@ -172,6 +173,7 @@ def _do_handshake(self):
         self._server_version = self._check_server_version(
             handshake["server_version_original"]
         )
+        CharacterSet.set_mysql_version(self._server_version)
 
         if not handshake["capabilities"] & ClientFlag.SSL:
             if self._auth_plugin == "mysql_clear_password":

Diff for: lib/mysql/connector/connection.py

@@ -133,6 +133,7 @@ def _do_handshake(self):
         self._server_version = self._check_server_version(
             self._handshake["server_version_original"]
         )
+        CharacterSet.set_mysql_version(self._server_version)
 
     @property
     def _server_status(self):

Diff for: lib/mysql/connector/connection_cext.py

@@ -32,7 +32,7 @@
 
 from abc import ABC, ABCMeta
 
-from .charsets import MYSQL_CHARACTER_SETS
+from .charsets import MYSQL_CHARACTER_SETS, MYSQL_CHARACTER_SETS_57
 from .errors import ProgrammingError
 
 MAX_PACKET_LENGTH = 16777215
@@ -678,10 +678,22 @@ class CharacterSet(_Constants):
     """
 
     desc = MYSQL_CHARACTER_SETS
+    mysql_version = (8, 0)
 
     # Multi-byte character sets which use 5c (backslash) in characters
     slash_charsets = (1, 13, 28, 84, 87, 88)
 
+    @classmethod
+    def set_mysql_version(cls, version):
+        """Set the MySQL major version and change the charset mapping if is 5.7.
+
+        Args:
+            version (tuple): MySQL version tuple.
+        """
+        cls.mysql_version = version[:2]
+        if cls.mysql_version == (5, 7):
+            cls.desc = MYSQL_CHARACTER_SETS_57
+
     @classmethod
     def get_info(cls, setid):
         """Retrieves character set information as tuple using an ID
@@ -757,6 +769,8 @@ def get_charset_info(cls, charset=None, collation=None):
             except IndexError:
                 ProgrammingError(f"Character set ID {charset} unknown")
 
+        if charset in ("utf8", "utf-8") and cls.mysql_version == (8, 0):
+            charset = "utf8mb4"
         if charset is not None and collation is None:
             info = cls.get_default_collation(charset)
             return (info[2], info[1], info[0])

Diff for: lib/mysql/connector/constants.py

@@ -63,7 +63,7 @@ def __init__(self, charset="utf8", use_unicode=True, str_fallback=False):
 
     def set_charset(self, charset):
         """Set character set"""
-        if charset == "utf8mb4":
+        if charset in ("utf8mb4", "utf8mb3"):
             charset = "utf8"
         if charset is not None:
             self.charset = charset