Note: bump-pydantic is still in beta

Replace its use with
`pydantic.BaseModel.model_dump(mode='json', exclude_none=True)`

Its use can be replaced with `pydantic.BaseModel.model_construct`

But mark it as deprecated

But mark it as deprecated

Specifying the most compact separators to use in JSON
serialization, `(",", ":")`, is no longer needed for
they are the default separators to use for JSON serialization
in Pydantic V2

…date_json()`

This is the implicit default in Pydantic V1

…lue`

…igest`

…rements`

…`BaseModel.model_json_schema`

`description` instead of `descriptions`

…struct` in tests

This commit modifies method signatures for
 `post_process_json_schema` by adding an additional
  parameter `model`. This parameter is also used in
  the method bodies to set the 'title' if it does not
  exist in the schema. These modifications are needed
  because Pydantic V2 doesn't provide the `title` in
  the JSON schema until the very end.

This commit changes the import path for Pydantic's
ModelMetaclass. The old path,
`pydantic.main.ModelMetaclass`, has been replaced by
the new path,
`pydantic._internal._model_construction.ModelMetaclass`.

The schema's definition path has changed, so this commit
updates the testing assertions accordingly. The key
 "definitions" has been replaced with "$defs" in Pydantic V2

The pydantic library's location for ModelMetaclass has
changed, so the paths had to be updated in the tests.

…enerate_context`

…est_duplicate_classes`

…te_classes`

…properties_mismatch`

Get outer type using `type.get_origin()` instead of by stringifying.

Co-authored-by: John T. Wodder II <[email protected]>

The update modifies the JSON serialization methods used
in the ZarrChecksumListing class. The code now uses `json.dumps`
 instead of previously used `BaseModel.model_dump_json()`.
 These changes ensure JSON serializations are done in the most
 compact manner by specifying the separators explicitly instead
 of depending on the default behavior of 'BaseModel.model_dump_json()'.

The new names are more appropriate per the use of the arguments

Use `inspect.isclass` and `issubclass` instead of relying on
internal metaclass

to modify JSON schema

Use `BaseModel.__get_pydantic_json_schema__()`
to modify JSON schema instead of
`BaseModel.model_config["json_schema_extra"]`.
This allows the automatically generated `title`
attribute to be available for modification.

Adding JSON schema to `Pydantic.ByteSize`
through annotation instead of subclassing
allows objects of type `Pydantic.ByteSize`
to assign to a `ByteSizeJsonSchema` attribute
without static type checker complain

Adding JSON schema to `Pydantic.ByteSize`
through annotation instead of subclassing
allows objects of type `Pydantic.ByteSize`
to assign to a `ByteSizeJsonSchema` attribute
without static type checker complain

This subclass can be used to overrides
the default JSON schema generation to match
that of the behavior of Pydantic V1.

Update the test so that it is adjusted to
the error representation in Pydantic V2

Update the test so that it is adjusted to
the error representation in Pydantic V2

Update the test so that it is adjusted to
the error representation in Pydantic V2

Update `metadata.validate` so that it is
adjusted to the error representation in
Pydantic V2

Unpacking of a dict as kwargs doesn't play
well with mypy. See the example below.
```
import json

ENCODING_KWARGS = {"separators": (",", ":")}

d = {"a": 1, "b": 2}
content = json.dumps(d, **ENCODING_KWARGS)
```

URL types in Pydantic V2 are no longer a
subclass of `str`. For uses that expect
a `str`, conversion to `str` is needed.

`field.json_schema_extra` can be either `None` or a callable as well

The values of `field.json_schema_extra` can
be one of many types

`BaseModel.model_construct()` is purposely
defined for use without conforming to
requirements specified in the model.

For `expected_errors` in `test_dantimeta_1`

Of `field.json_schema_extra` in
`test_properties_mismatch()`

Remove unreachable code block in
`test_duplicate_classes`. According to
Satra, this block must be outdated.

This solution is kept for now so that context generator
can maintain the same behavior

…ave incompatible changes)

for more information, see https://pre-commit.ci

The logic of the result is equivalent to
the one before the modification

There is no change in the data model
because of the Pydantic upgrade to V2

For dandi JSON schema 0.7.0 or above

So that mypy doesn't complain

0.6.5 is more sensible since the changes
in the schema are not breaking changes,
and there is no change in the models and
the JSON serialization of these models