* Refactored to avoid symbol conflicts
* Forgot a replacement
* Found a symbol
* Undid symbol replacement for PyInit__cmessage
* Changed cbson too
Co-authored-by: Ben Warner <ben.warner@mongodb.com>
Fix DeprecationWarning: PY_SSIZE_T_CLEAN will be required for '#' formats
Fix DeprecationWarning: isAlive() is deprecated, use is_alive() instead
Fix SyntaxWarning: invalid escape sequence
Test Python 3.8 on Travis
Fix batched op_msg/op_query reference leak of overflow doc.
Fix theoretically possible (but practically impossible) reference leak of
$clusterTime in op_query.
Optimization: Don't encode document past the batch size in batched op_query.
Replace OP_INSERT/OP_UPDATE/OP_DELETE usage with OP_MSG with moreToCome
flag when the server is 3.6 or greater.
Unack'ed OP_MSG supports bypass_document_validation.
getLastError updatedExisting is always false on mongos after an OP_MSG
unack'ed update.
Ordered op_insert should check for errors on the last insert.
Use _Bulk API for Collection.insert.
Avoid sending acknowledged inserts as much as possible.
max_cmd_size already accounts for trailing null bytes.
This change adds monitoring of bulk write operations (i.e.
Collection.bulk_write, Collection.insert_many, Collection.insert with multiple
documents, Bulk.execute, etc.). It also fixes bugs in conversion of legacy
write results to write command result format and conversion of legacy queries
to find command documents. Finally, it adds an operation_id attribute to the
published events to tie related events together.
As with the Python functions, the C implementations of _do_batched_insert and
_do_batched_write command must now take a SocketInfo, not a MongoClient, to
ensure they send the OP_INSERT or write command over the same socket used to
determine the server's wire protocol version.
This change resolves four issues:
PYTHON-826 The new codec_options submodule is moved from pymongo to bson.
PYTHON-827 Use codec_options in BSON APIs.
Functions and methods of the bson module that accepted the options as_class,
tz_aware, and uuid_subtype now accept a codec_options parameter instead.
For example, the function definition for bson.decode_all changes from this:
def decode_all(data, as_class=dict, tz_aware=True,
uuid_subtype=OLD_UUID_SUBTYPE)
to:
def decode_all(data, codec_options=CodecOptions())
The following functions are changed:
- decode_all
- decode_iter
- decode_file_iter
The following methods are changed:
- BSON.encode
- BSON.decode
This is a breaking change for any application that uses the BSON API directly
and changes any of the named parameter defaults. No changes are required for
applications that use the default values for these options. The behavior
remains the same.
PYTHON-828 Internal BSON module changes to support CodecOptions
The pure Python BSON module passes around a CodecOptions instance instead of
as_class, tz_aware, and uuid_subtype. C extensions pass these values around in
a struct.
PYTHON-801 Rename uuid_subtype to uuid_representation.
In PyMongo 2.6, message.insert was replaced with _do_batched insert. It was
deprecated but not removed in case third-party libraries might rely on it. In
PyMongo 2.7 we implemented the Bulk API (PYTHON-630) and message.insert found a
new use.
The idea here is to unify the handling of oversize documents when using
the bulk API in MongoDB 2.6 and previous versions. This also means that
using bulk Collection.insert against legacy servers will attempt to insert
all documents previous to the oversize document before raising.
A few changes:
- Eliminate the ordered parameter - no reason this needs to be a
separate param.
- Use integers to represent the operation instead of strings - this
will make part of the bulk API implementation a bit cleaner and
avoid a bunch of string comparisons.
- Clean up namespace and collection name handling in collection.py - No
need to create a bunch of new strings or do unnecessary string copies.
- Pass client.disconnect to _check_command_response to deal with
"not master".
