PYTHON-6040 Preserve version-to-name mapping in Client Metadata

doc/changelog.rst

@@ -10,8 +10,23 @@ Bug fixes
 - Fixed a bug where the synchronous client could permanently deadlock under
   gevent when a greenlet was killed while checking a connection back into
   the pool (`PYTHON-6074`_).
+- ``MongoClient.append_metadata()`` and ``AsyncMongoClient.append_metadata()``
+  now detect duplicates by comparing the whole
+  :class:`~pymongo.driver_info.DriverInfo` instead of only its name. The
+  comparison is exact, so drivers that differ in name case or platform are no
+  longer treated as duplicates (`PYTHON-6040`_).
+- ``driver.name`` and ``driver.version`` in the handshake metadata are now
+  ``|``-delimited lists with 1:1 index correspondence, including empty version
+  entries for the built-in ``|c`` and ``|async`` name segments
+  (`PYTHON-6040`_).
+- Fixed a bug where truncating the handshake metadata to 512 bytes could leave
+  ``driver.name`` and ``driver.version`` with different numbers of ``|``
+  delimiters (`PYTHON-6040`_).
+- :class:`~pymongo.driver_info.DriverInfo` now raises :class:`ValueError` when
+  any field contains the reserved ``|`` delimiter (`PYTHON-6040`_).
 
 .. _PYTHON-6074: https://jira.mongodb.org/browse/PYTHON-6074
+.. _PYTHON-6040: https://jira.mongodb.org/browse/PYTHON-6040
 
 Changes in Version 4.18.1 (2026/09/10)
 --------------------------------------

pymongo/driver_info.py

@@ -29,8 +29,14 @@ class DriverInfo(namedtuple("DriverInfo", ["name", "version", "platform"])):
     The MongoDB server logs PyMongo's name, version, and platform whenever
     PyMongo establishes a connection. A driver implemented on top of PyMongo
     can add its own info to this log message. Initialize with three strings
-    like 'MyDriver', '1.2.3', 'some platform info'. Any of these strings may be
-    None to accept PyMongo's default.
+    like 'MyDriver', '1.2.3', 'some platform info'. Any of these strings may
+    be None. A None or empty name or version appends an empty metadata
+    segment, keeping the ``driver.name`` and ``driver.version`` segment
+    counts aligned. A None or empty platform omits the platform.
+
+    The ``|`` character is the reserved delimiter used to join appended
+    metadata, so it must not appear in any of the fields. A
+    :class:`ValueError` is raised if it does.
     """
 
     def __new__(
@@ -42,5 +48,7 @@ def __new__(
                 raise TypeError(
                     f"Wrong type for DriverInfo {key} option, value must be an instance of str, not {type(value)}"
                 )
+            if value and "|" in value:
+                raise ValueError(f"DriverInfo {key} must not contain the '|' delimiter")
 
         return self

pymongo/pool_options.py

@@ -24,6 +24,7 @@
 import platform
 import sys
 from collections.abc import MutableMapping
+from contextlib import AbstractContextManager, nullcontext
 from pathlib import Path
 from typing import TYPE_CHECKING, Any, Optional
 
@@ -37,6 +38,7 @@
     WAIT_QUEUE_TIMEOUT,
     has_c,
 )
+from pymongo.lock import _create_lock
 
 if TYPE_CHECKING:
     from pymongo.auth_shared import MongoCredential
@@ -200,6 +202,25 @@ def _metadata_env() -> dict[str, Any]:
 _MAX_METADATA_SIZE = 512
 
 
+def _truncate_utf8(content: str, overflow: int) -> str:
+    """Trim `overflow` UTF-8 bytes from the end of content, keeping a valid prefix."""
+    if overflow <= 0:
+        return content
+    data = content.encode("utf-8")
+    if len(data) <= overflow:
+        return ""
+    return data[: len(data) - overflow].decode("utf-8", errors="ignore")
+
+
+def _normalize_driver(driver: DriverInfo) -> DriverInfo:
+    """Treat None and "" as equivalent unset fields for deduplication."""
+    return driver._replace(
+        name=driver.name or "",
+        version=driver.version or "",
+        platform=driver.platform or "",
+    )
+
+
 # See: https://github.com/mongodb/specifications/blob/master/source/mongodb-handshake/handshake.md#limitations
 def _truncate_metadata(metadata: MutableMapping[str, Any]) -> None:
     """Perform metadata truncation."""
@@ -226,34 +247,31 @@ def _truncate_metadata(metadata: MutableMapping[str, Any]) -> None:
     overflow = encoded_size - _MAX_METADATA_SIZE
     plat = metadata.get("platform", "")
     if plat:
-        plat = plat[:-overflow]
+        plat = _truncate_utf8(plat, overflow)
     if plat:
         metadata["platform"] = plat
     else:
         metadata.pop("platform", None)
-    encoded_size = len(bson.encode(metadata))
-    if encoded_size <= _MAX_METADATA_SIZE:
-        return
-    # 5. Truncate driver info.
-    overflow = encoded_size - _MAX_METADATA_SIZE
+    # 5. Truncate driver info, keeping name and version 1:1 index-aligned.
     driver = metadata.get("driver", {})
     if driver:
-        # Truncate driver version.
-        driver_version = driver.get("version")[:-overflow]
-        if len(driver_version) >= len(_METADATA["driver"]["version"]):
-            metadata["driver"]["version"] = driver_version
-        else:
-            metadata["driver"]["version"] = _METADATA["driver"]["version"]
-        encoded_size = len(bson.encode(metadata))
-        if encoded_size <= _MAX_METADATA_SIZE:
-            return
-        # Truncate driver name.
-        overflow = encoded_size - _MAX_METADATA_SIZE
-        driver_name = driver.get("name")[:-overflow]
-        if len(driver_name) >= len(_METADATA["driver"]["name"]):
-            metadata["driver"]["name"] = driver_name
-        else:
-            metadata["driver"]["name"] = _METADATA["driver"]["name"]
+        # Keep the name and version segments paired so they stay 1:1 aligned,
+        # trimming wrapper content and dropping paired segments only as a
+        # last resort.
+        pairs = list(zip(driver.get("name", "").split("|"), driver.get("version", "").split("|")))
+        while True:
+            driver["name"] = "|".join(name for name, _ in pairs)
+            driver["version"] = "|".join(version for _, version in pairs)
+            overflow = len(bson.encode(metadata)) - _MAX_METADATA_SIZE
+            if overflow <= 0 or len(pairs) <= 1:
+                break
+            last_name, last_version = pairs[-1]
+            if last_version:
+                pairs[-1] = (last_name, _truncate_utf8(last_version, overflow))
+            elif last_name:
+                pairs[-1] = (_truncate_utf8(last_name, overflow), last_version)
+            else:
+                pairs.pop()
 
 
 # If the first getaddrinfo call of this interpreter's life is on a thread,
@@ -277,6 +295,7 @@ class PoolOptions:
     """
 
     __slots__ = (
+        "__appended_drivers",
         "__appname",
         "__compression_settings",
         "__connect_timeout",
@@ -288,6 +307,7 @@ class PoolOptions:
         "__max_idle_time_seconds",
         "__max_pool_size",
         "__metadata",
+        "__metadata_lock",
         "__min_pool_size",
         "__pause_enabled",
         "__server_api",
@@ -336,6 +356,11 @@ def __init__(
         self.__load_balanced = load_balanced
         self.__credentials = credentials
         self.__metadata = copy.deepcopy(_METADATA)
+        self.__appended_drivers: list[DriverInfo] = []
+        # Only the synchronous client can append metadata from multiple threads.
+        self.__metadata_lock: AbstractContextManager[bool | None] = (
+            _create_lock() if is_sync else nullcontext()
+        )
 
         if appname:
             self.__metadata["application"] = {"name": appname}
@@ -353,11 +378,19 @@ def __init__(
                 self.__metadata["driver"]["name"],
                 "c",
             )
+            self.__metadata["driver"]["version"] = "{}|{}".format(
+                self.__metadata["driver"]["version"],
+                "",
+            )
         if not is_sync:
             self.__metadata["driver"]["name"] = "{}|{}".format(
                 self.__metadata["driver"]["name"],
                 "async",
             )
+            self.__metadata["driver"]["version"] = "{}|{}".format(
+                self.__metadata["driver"]["version"],
+                "",
+            )
         if driver:
             self._update_metadata(driver)
 
@@ -368,28 +401,38 @@ def __init__(
         _truncate_metadata(self.__metadata)
 
     def _update_metadata(self, driver: DriverInfo) -> None:
-        """Updates the client's metadata"""
-        if driver.name and driver.name.lower() in self.__metadata["driver"]["name"].lower().split(
-            "|"
-        ):
-            return
-
-        metadata = copy.deepcopy(self.__metadata)
-
-        if driver.name:
-            metadata["driver"]["name"] = "{}|{}".format(
-                metadata["driver"]["name"],
-                driver.name,
-            )
-        if driver.version:
+        """Updates the client's metadata."""
+        with self.__metadata_lock:
+            driver = _normalize_driver(driver)
+            if driver in self.__appended_drivers:
+                return
+
+            name_delims = self.__metadata["driver"]["name"].count("|")
+            version_delims = self.__metadata["driver"]["version"].count("|")
+            metadata = copy.deepcopy(self.__metadata)
+
+            metadata["driver"]["name"] = "{}|{}".format(metadata["driver"]["name"], driver.name)
             metadata["driver"]["version"] = "{}|{}".format(
-                metadata["driver"]["version"],
-                driver.version,
+                metadata["driver"]["version"], driver.version
             )
-        if driver.platform:
-            metadata["platform"] = "{}|{}".format(metadata["platform"], driver.platform)
-
-        self.__metadata = metadata
+            if driver.platform:
+                if "platform" in metadata:
+                    metadata["platform"] = "{}|{}".format(metadata["platform"], driver.platform)
+                else:
+                    metadata["platform"] = driver.platform
+
+            _truncate_metadata(metadata)
+
+            self.__metadata = metadata
+
+            # Only track drivers whose appended name/version pair survived
+            # truncation (i.e. both gained a segment), so __appended_drivers
+            # stays bounded and the dedup membership check stays fast.
+            if (
+                metadata["driver"]["name"].count("|") > name_delims
+                and metadata["driver"]["version"].count("|") > version_delims
+            ):
+                self.__appended_drivers.append(driver)
 
     @property
     def _credentials(self) -> Optional[MongoCredential]: