Bases: Middleware
Middleware that sorts documents by integer doc_id on write.
Wraps a Storage and, on each write,
converts the stringified document ids back to integers and forces
the underlying storage to sort keys, so documents are serialized in
numeric doc_id order rather than lexicographic string order
(where "10" would sort before "2").
Warning
This middleware may break storages that cannot serialize integer
keys, since it passes integer doc_id values through to the
underlying storage.
Source code in src/tinydantic/tinydb/middlewares.py
| class SortIntDocIDsMiddleware(Middleware):
"""Middleware that sorts documents by integer ``doc_id`` on write.
Wraps a [Storage][tinydb.storages.Storage] and, on each write,
converts the stringified document ids back to integers and forces
the underlying storage to sort keys, so documents are serialized in
numeric ``doc_id`` order rather than lexicographic string order
(where ``"10"`` would sort before ``"2"``).
Warning:
This middleware may break storages that cannot serialize integer
keys, since it passes integer ``doc_id`` values through to the
underlying storage.
"""
def __init__(self, storage_cls: type[Storage]) -> None:
"""Wrap ``storage_cls`` with integer-``doc_id`` sorting.
Args:
storage_cls: The storage class to wrap.
"""
super().__init__(storage_cls)
def write(self, data: dict[str, dict[str, Any]]) -> None:
"""Write ``data`` with documents sorted by integer ``doc_id``.
Args:
data: The table data to write, keyed by table name then
document id.
"""
# 3-step "trick" to numerically sort the documents in each table
# by their integer doc_id:
# 1. Dict comprehension to convert doc_id from strings to
# integers.
#
# Note: this is required even though the type of doc_id is
# integer because doc_id is preemptively converted to a
# string before being passed to the middleware/storage class
# (see https://github.com/msiemens/tinydb/discussions/466).
int_keys_data: dict[str, dict[int | str, Any]] = {}
for table, table_data in data.items():
int_keys_data[table] = {
int(doc_id): value for doc_id, value in table_data.items()
}
# 2. Force the storage class to sort by key on write.
#
# For example:
# json.dumps(data, fp, sort_keys=True, ...)
# yaml.dump(data, fp, sort_keys=True, ...)
# etc...
#
# Note: this requires that the storage class supports the
# sort_keys keyword argument to dump, and it overwrites the
# value specified when the storage class was initialized.
self.storage.kwargs["sort_keys"] = True # type: ignore[attr-defined]
# 3. Instruct the storage class to write the data using integer
# keys. This works for JSONStorage because json.dumps() will
# coerce integer document IDs to strings (JSON requires that
# keys are strings). It also works for YAMLStorage because
# the YAML spec allows integer keys.
#
# TinyDB's Storage.write() expects data to be of type
# dict[str, dict[str, Any]] but we're passing in data of type
# dict[str, dict[int, Any]] instead.
#
# As a result, we need to tell the type checker to ignore
# arg-type type errors.
self.storage.write(data=int_keys_data) # type: ignore[arg-type]
|
Wrap storage_cls with integer-doc_id sorting.
Parameters:
| Name |
Type |
Description |
Default |
storage_cls
|
type[Storage]
|
The storage class to wrap.
|
required
|
Source code in src/tinydantic/tinydb/middlewares.py
| def __init__(self, storage_cls: type[Storage]) -> None:
"""Wrap ``storage_cls`` with integer-``doc_id`` sorting.
Args:
storage_cls: The storage class to wrap.
"""
super().__init__(storage_cls)
|
Write data with documents sorted by integer doc_id.
Parameters:
| Name |
Type |
Description |
Default |
data
|
dict[str, dict[str, Any]]
|
The table data to write, keyed by table name then
document id.
|
required
|
Source code in src/tinydantic/tinydb/middlewares.py
| def write(self, data: dict[str, dict[str, Any]]) -> None:
"""Write ``data`` with documents sorted by integer ``doc_id``.
Args:
data: The table data to write, keyed by table name then
document id.
"""
# 3-step "trick" to numerically sort the documents in each table
# by their integer doc_id:
# 1. Dict comprehension to convert doc_id from strings to
# integers.
#
# Note: this is required even though the type of doc_id is
# integer because doc_id is preemptively converted to a
# string before being passed to the middleware/storage class
# (see https://github.com/msiemens/tinydb/discussions/466).
int_keys_data: dict[str, dict[int | str, Any]] = {}
for table, table_data in data.items():
int_keys_data[table] = {
int(doc_id): value for doc_id, value in table_data.items()
}
# 2. Force the storage class to sort by key on write.
#
# For example:
# json.dumps(data, fp, sort_keys=True, ...)
# yaml.dump(data, fp, sort_keys=True, ...)
# etc...
#
# Note: this requires that the storage class supports the
# sort_keys keyword argument to dump, and it overwrites the
# value specified when the storage class was initialized.
self.storage.kwargs["sort_keys"] = True # type: ignore[attr-defined]
# 3. Instruct the storage class to write the data using integer
# keys. This works for JSONStorage because json.dumps() will
# coerce integer document IDs to strings (JSON requires that
# keys are strings). It also works for YAMLStorage because
# the YAML spec allows integer keys.
#
# TinyDB's Storage.write() expects data to be of type
# dict[str, dict[str, Any]] but we're passing in data of type
# dict[str, dict[int, Any]] instead.
#
# As a result, we need to tell the type checker to ignore
# arg-type type errors.
self.storage.write(data=int_keys_data) # type: ignore[arg-type]
|