PYTHON-3703 Fix typos and add codespell pre-commit hook (#1203)

Update pre-commit config adding "codespell" hook

Author: jcfr

.pre-commit-config.yaml

@@ -83,3 +83,16 @@ repos:
     files: \.py$
     exclude: "^(test|tools)/"
     stages: [manual]
+
+- repo: https://github.com/codespell-project/codespell
+  rev: "v2.2.4"
+  hooks:
+  - id: codespell
+    # Examples of errors or updates to justify the exceptions:
+    # - test/test_on_demand_csfle.py:44: FLE ==> FILE
+    # - test/test_bson.py:1043: fo ==> of, for, to, do, go
+    # - test/bson_corpus/decimal128-4.json:98: Infinit ==> Infinite
+    # - test/test_bson.py:267: isnt ==> isn't
+    # - test/versioned-api/crud-api-version-1-strict.json:514: nin ==> inn, min, bin, nine
+    # - test/test_client.py:188: te ==> the, be, we, to
+    args: ["-L", "fle,fo,infinit,isnt,nin,te"]

bson/__init__.py

@@ -1282,7 +1282,7 @@ def decode_file_iter(
         # Read size of next object.
         size_data = file_obj.read(4)
         if not size_data:
-            break  # Finished with file normaly.
+            break  # Finished with file normally.
         elif len(size_data) != 4:
             raise InvalidBSON("cut off in middle of objsize")
         obj_size = _UNPACK_INT_FROM(size_data, 0)[0] - 4

bson/_cbsonmodule.c

@@ -968,7 +968,7 @@ static int _write_element_to_buffer(PyObject* self, buffer_t buffer,
         }
     }
 
-    /* No _type_marker attibute or not one of our types. */
+    /* No _type_marker attribute or not one of our types. */
 
     if (PyBool_Check(value)) {
         const char c = (value == Py_True) ? 0x01 : 0x00;

bson/objectid.py

@@ -231,7 +231,7 @@ def __getstate__(self) -> bytes:
 
     def __setstate__(self, value: Any) -> None:
         """explicit state set from pickling"""
-        # Provide backwards compatability with OIDs
+        # Provide backwards compatibility with OIDs
         # pickled with pymongo-1.9 or older.
         if isinstance(value, dict):
             oid = value["_ObjectId__id"]

bson/time64.c

@@ -73,7 +73,7 @@ static const Year     years_in_gregorian_cycle   = 400;
 #define               days_in_gregorian_cycle      ((365 * 400) + 100 - 4 + 1)
 static const Time64_T seconds_in_gregorian_cycle = days_in_gregorian_cycle * 60LL * 60LL * 24LL;
 
-/* Year range we can trust the time funcitons with */
+/* Year range we can trust the time functions with */
 #define MAX_SAFE_YEAR 2037
 #define MIN_SAFE_YEAR 1971
 
@@ -739,7 +739,7 @@ struct TM *cbson_localtime64_r (const Time64_T *time, struct TM *local_tm)
     /* GMT is Jan 1st, xx01 year, but localtime is still Dec 31st
        in a non-leap xx00.  There is one point in the cycle
        we can't account for which the safe xx00 year is a leap
-       year.  So we need to correct for Dec 31st comming out as
+       year.  So we need to correct for Dec 31st coming out as
        the 366th day of the year.
     */
     if( !IS_LEAP(local_tm->tm_year) && local_tm->tm_yday == 365 )

doc/changelog.rst

@@ -2805,7 +2805,7 @@ Important New Features:
 - The URI parser has been moved into its own module and can be used directly
   by application code.
 - AutoReconnect exception now provides information about the error that
-  actually occured instead of a generic failure message.
+  actually occurred instead of a generic failure message.
 - A number of new helper methods have been added with options for setting and
   unsetting cursor flags, re-indexing a collection, fsync and locking a server,
   and getting the server's current operations.
@@ -2930,7 +2930,7 @@ Issues resolved
 - `PYTHON-186 <https://jira.mongodb.org/browse/PYTHON-186>`_:
   When storing integers, type is selected according to value instead of type
 - `PYTHON-173 <https://jira.mongodb.org/browse/PYTHON-173>`_:
-  as_class option is not propogated by Cursor.clone
+  as_class option is not propagated by Cursor.clone
 - `PYTHON-113 <https://jira.mongodb.org/browse/PYTHON-113>`_:
   Redunducy in MasterSlaveConnection
 

doc/contributors.rst

@@ -93,3 +93,4 @@ The following is a list of people who have contributed to
 - Ishmum Jawad Khan (ishmum123)
 - Arie Bovenberg (ariebovenberg)
 - Ben Warner (bcwarner)
+- Jean-Christophe Fillion-Robin (jcfr)

doc/examples/bulk.rst

@@ -80,7 +80,7 @@ of operations performed.
 
 The first write failure that occurs (e.g. duplicate key error) aborts the
 remaining operations, and PyMongo raises
-:class:`~pymongo.errors.BulkWriteError`. The :attr:`details` attibute of
+:class:`~pymongo.errors.BulkWriteError`. The :attr:`details` attribute of
 the exception instance provides the execution results up until the failure
 occurred and details about the failure - including the operation that caused
 the failure.

doc/examples/type_hints.rst

@@ -264,7 +264,7 @@ Troubleshooting
 
 Client Type Annotation
 ~~~~~~~~~~~~~~~~~~~~~~
-If you forget to add a type annotation for a :class:`~pymongo.mongo_client.MongoClient` object you may get the followig ``mypy`` error::
+If you forget to add a type annotation for a :class:`~pymongo.mongo_client.MongoClient` object you may get the following ``mypy`` error::
 
   from pymongo import MongoClient
   client = MongoClient()  # error: Need type annotation for "client"
@@ -313,10 +313,10 @@ Another example is trying to set a value on a :class:`~bson.raw_bson.RawBSONDocu
     coll = client.test.test
     doc = {"my": "doc"}
     coll.insert_one(doc)
-    retreived = coll.find_one({"_id": doc["_id"]})
-    assert retreived is not None
-    assert len(retreived.raw) > 0
-    retreived[
+    retrieved = coll.find_one({"_id": doc["_id"]})
+    assert retrieved is not None
+    assert len(retrieved.raw) > 0
+    retrieved[
         "foo"
     ] = "bar"  # error: Unsupported target for indexed assignment
                # ("RawBSONDocument")  [index]

gridfs/grid_file.py

@@ -381,7 +381,7 @@ def write(self, data: Any) -> None:
     def writelines(self, sequence: Iterable[Any]) -> None:
         """Write a sequence of strings to the file.
 
-        Does not add seperators.
+        Does not add separators.
         """
         for line in sequence:
             self.write(line)

pymongo/collection.py

@@ -2076,7 +2076,7 @@ def drop_indexes(
     ) -> None:
         """Drops all indexes on this collection.
 
-        Can be used on non-existant collections or collections with no indexes.
+        Can be used on non-existent collections or collections with no indexes.
         Raises OperationFailure on an error.
 
         :Parameters:
@@ -2112,7 +2112,7 @@ def drop_index(
     ) -> None:
         """Drops the specified index on this collection.
 
-        Can be used on non-existant collections or collections with no
+        Can be used on non-existent collections or collections with no
         indexes.  Raises OperationFailure on an error (e.g. trying to
         drop an index that does not exist). `index_or_name`
         can be either an index name (as returned by `create_index`),
@@ -2683,7 +2683,7 @@ def rename(
         if not new_name or ".." in new_name:
             raise InvalidName("collection names cannot be empty")
         if new_name[0] == "." or new_name[-1] == ".":
-            raise InvalidName("collecion names must not start or end with '.'")
+            raise InvalidName("collection names must not start or end with '.'")
         if "$" in new_name and not new_name.startswith("oplog.$main"):
             raise InvalidName("collection names must not contain '$'")
 

pymongo/common.py

@@ -469,7 +469,7 @@ def validate_document_class(
         raise TypeError(
             "%s must be dict, bson.son.SON, "
             "bson.raw_bson.RawBSONDocument, or a "
-            "sublass of collections.MutableMapping" % (option,)
+            "subclass of collections.MutableMapping" % (option,)
         )
     return value
 

pymongo/message.py

@@ -1077,7 +1077,7 @@ def _batched_op_msg_impl(operation, command, docs, ack, opts, ctx, buf):
         new_message_size = buf.tell() + doc_length
         # Does first document exceed max_message_size?
         doc_too_large = idx == 0 and (new_message_size > max_message_size)
-        # When OP_MSG is used unacknowleged we have to check
+        # When OP_MSG is used unacknowledged we have to check
         # document size client side or applications won't be notified.
         # Otherwise we let the server deal with documents that are too large
         # since ordered=False causes those documents to be skipped instead of

pymongo/ocsp_support.py

@@ -312,7 +312,7 @@ def _ocsp_callback(conn, ocsp_bytes, user_data):
             return 0
         if not user_data.check_ocsp_endpoint:
             _LOGGER.debug("OCSP endpoint checking is disabled, soft fail.")
-            # No stapled OCSP response, checking responder URI diabled, soft fail.
+            # No stapled OCSP response, checking responder URI disabled, soft fail.
             return 1
         # https://tools.ietf.org/html/rfc6960#section-3.1
         ext = _get_extension(cert, _AuthorityInformationAccess)

pymongo/results.py

@@ -45,7 +45,7 @@ def acknowledged(self) -> bool:
 
         .. note::
           If the :attr:`acknowledged` attribute is ``False`` all other
-          attibutes of this class will raise
+          attributes of this class will raise
           :class:`~pymongo.errors.InvalidOperation` when accessed. Values for
           other attributes cannot be determined if the write operation was
           unacknowledged.

pymongo/uri_parser.py

@@ -42,7 +42,7 @@
 def _unquoted_percent(s):
     """Check for unescaped percent signs.
 
-    :Paramaters:
+    :Parameters:
         - `s`: A string. `s` can have things like '%25', '%2525',
            and '%E2%85%A8' but cannot have unquoted percent like '%foo'.
     """
@@ -64,7 +64,7 @@ def parse_userinfo(userinfo: str) -> Tuple[str, str]:
     Returns a 2-tuple containing the unescaped username followed
     by the unescaped password.
 
-    :Paramaters:
+    :Parameters:
         - `userinfo`: A string of the form <username>:<password>
     """
     if "@" in userinfo or userinfo.count(":") > 1 or _unquoted_percent(userinfo):

test/client-side-encryption/spec/legacy/fle2v2-MissingKey.json

@@ -55,7 +55,7 @@
   "key_vault_data": [],
   "tests": [
     {
-      "description": "FLE2 encrypt fails with mising key",
+      "description": "FLE2 encrypt fails with missing key",
       "clientOptions": {
         "autoEncryptOpts": {
           "kmsProviders": {
@@ -86,7 +86,7 @@
       ]
     },
     {
-      "description": "FLE2 decrypt fails with mising key",
+      "description": "FLE2 decrypt fails with missing key",
       "clientOptions": {
         "autoEncryptOpts": {
           "kmsProviders": {

test/csot/deprecated-options.json

@@ -1,5 +1,5 @@
 {
-  "description": "operations ignore deprected timeout options if timeoutMS is set",
+  "description": "operations ignore deprecated timeout options if timeoutMS is set",
   "schemaVersion": "1.9",
   "runOnRequirements": [
     {

test/mod_wsgi_test/mod_wsgi_test.conf

@@ -27,7 +27,7 @@ WSGISocketPrefix /tmp/
 
     WSGIProcessGroup mod_wsgi_test
 
-    # For the convienience of unittests, rather than hard-code the location of
+    # For the convenience of unittests, rather than hard-code the location of
     # mod_wsgi_test.wsgi, include it in the URL, so
     # http://localhost/location-of-pymongo-checkout will work:
 

test/mypy_fails/raw_bson_document.py

@@ -5,9 +5,9 @@
 coll = client.test.test
 doc = {"my": "doc"}
 coll.insert_one(doc)
-retreived = coll.find_one({"_id": doc["_id"]})
-assert retreived is not None
-assert len(retreived.raw) > 0
-retreived[
+retrieved = coll.find_one({"_id": doc["_id"]})
+assert retrieved is not None
+assert len(retrieved.raw) > 0
+retrieved[
     "foo"
 ] = "bar"  # error: Unsupported target for indexed assignment ("RawBSONDocument")  [index]

test/mypy_fails/typedict_client.py

@@ -10,9 +10,9 @@ class Movie(TypedDict):
 
 client: MongoClient[Movie] = MongoClient()
 coll = client.test.test
-retreived = coll.find_one({"_id": "foo"})
-assert retreived is not None
-assert retreived["year"] == 1
+retrieved = coll.find_one({"_id": "foo"})
+assert retrieved is not None
+assert retrieved["year"] == 1
 assert (
-    retreived["name"] == 2
+    retrieved["name"] == 2
 )  # error: Non-overlapping equality check (left operand type: "str", right operand type: "Literal[2]")  [comparison-overlap]

test/test_read_write_concern_spec.py

@@ -280,7 +280,7 @@ def run_test(self):
                 self.assertEqual(write_concern.acknowledged, test_case["isAcknowledged"])
                 self.assertEqual(write_concern.is_server_default, test_case["isServerDefault"])
         if "readConcern" in test_case:
-            # Any string for 'level' is equaly valid
+            # Any string for 'level' is equally valid
             read_concern = ReadConcern(**test_case["readConcern"])
             self.assertEqual(read_concern.document, test_case["readConcernDocument"])
             self.assertEqual(not bool(read_concern.level), test_case["isServerDefault"])

test/test_typing.py

@@ -111,11 +111,11 @@ def test_insert_find(self) -> None:
         coll2 = self.client.test.test2
         result = self.coll.insert_one(doc)
         self.assertEqual(result.inserted_id, doc["_id"])
-        retreived = self.coll.find_one({"_id": doc["_id"]})
-        if retreived:
+        retrieved = self.coll.find_one({"_id": doc["_id"]})
+        if retrieved:
             # Documents returned from find are mutable.
-            retreived["new_field"] = 1
-            result2 = coll2.insert_one(retreived)
+            retrieved["new_field"] = 1
+            result2 = coll2.insert_one(retrieved)
             self.assertEqual(result2.inserted_id, result.inserted_id)
 
     def test_cursor_iterable(self) -> None:
@@ -182,9 +182,9 @@ def test_default_document_type(self) -> None:
         coll = client.test.test
         doc = {"my": "doc"}
         coll.insert_one(doc)
-        retreived = coll.find_one({"_id": doc["_id"]})
-        assert retreived is not None
-        retreived["a"] = 1
+        retrieved = coll.find_one({"_id": doc["_id"]})
+        assert retrieved is not None
+        retrieved["a"] = 1
 
     def test_aggregate_pipeline(self) -> None:
         coll3 = self.client.test.test3
@@ -329,26 +329,26 @@ class TestDocumentType(unittest.TestCase):
     def test_default(self) -> None:
         client: MongoClient = MongoClient()
         coll = client.test.test
-        retreived = coll.find_one({"_id": "foo"})
-        assert retreived is not None
-        retreived["a"] = 1
+        retrieved = coll.find_one({"_id": "foo"})
+        assert retrieved is not None
+        retrieved["a"] = 1
 
     @only_type_check
     def test_explicit_document_type(self) -> None:
         client: MongoClient[Dict[str, Any]] = MongoClient()
         coll = client.test.test
-        retreived = coll.find_one({"_id": "foo"})
-        assert retreived is not None
-        retreived["a"] = 1
+        retrieved = coll.find_one({"_id": "foo"})
+        assert retrieved is not None
+        retrieved["a"] = 1
 
     @only_type_check
     def test_typeddict_document_type(self) -> None:
         client: MongoClient[Movie] = MongoClient()
         coll = client.test.test
-        retreived = coll.find_one({"_id": "foo"})
-        assert retreived is not None
-        assert retreived["year"] == 1
-        assert retreived["name"] == "a"
+        retrieved = coll.find_one({"_id": "foo"})
+        assert retrieved is not None
+        assert retrieved["year"] == 1
+        assert retrieved["name"] == "a"
 
     @only_type_check
     def test_typeddict_document_type_insertion(self) -> None:
@@ -450,17 +450,17 @@ def test_typeddict_find_notrequired(self):
     def test_raw_bson_document_type(self) -> None:
         client = MongoClient(document_class=RawBSONDocument)
         coll = client.test.test
-        retreived = coll.find_one({"_id": "foo"})
-        assert retreived is not None
-        assert len(retreived.raw) > 0
+        retrieved = coll.find_one({"_id": "foo"})
+        assert retrieved is not None
+        assert len(retrieved.raw) > 0
 
     @only_type_check
     def test_son_document_type(self) -> None:
         client = MongoClient(document_class=SON[str, Any])
         coll = client.test.test
-        retreived = coll.find_one({"_id": "foo"})
-        assert retreived is not None
-        retreived["a"] = 1
+        retrieved = coll.find_one({"_id": "foo"})
+        assert retrieved is not None
+        retrieved["a"] = 1
 
     def test_son_document_type_runtime(self) -> None:
         client = MongoClient(document_class=SON[str, Any], connect=False)