"""Storage of annotations.
This module contains a collection of classes for handling storage of
annotations in memory in addition to serialization/deserialization to/from
disk.
Definitions
-----------
For the sake of clarity it is helpful to define a few terms used throughout
this documentation.
Annotation
A geometry and associated properties.
Geometry
One of: a point, a polygon, or a line string.
.. figure:: ../images/geometries.png
:width: 512
Properties
Key-value pairs associated with a geometry.
"""
from __future__ import annotations
import contextlib
import copy
import io
import json
import os
import pickle
import sqlite3
import struct
import sys
import tempfile
import threading
import uuid
import zlib
from abc import ABC, abstractmethod
from collections import defaultdict
from collections.abc import Generator, Iterable, Iterator, MutableMapping
from dataclasses import dataclass, field
from functools import lru_cache
from pathlib import Path
from typing import (
IO,
TYPE_CHECKING,
Any,
Callable,
ClassVar,
)
import numpy as np
import pandas as pd
import shapely
from shapely import wkb as shapely_wkb
from shapely import wkt as shapely_wkt
from shapely.affinity import scale, translate
from shapely.geometry import LineString, Point, Polygon
from shapely.geometry import mapping as geometry2feature
from shapely.geometry import shape as feature2geometry
import tiatoolbox
from tiatoolbox import DuplicateFilter, logger
from tiatoolbox.annotation.dsl import (
PY_GLOBALS,
SQL_GLOBALS,
json_contains,
json_list_sum,
py_regexp,
)
from tiatoolbox.enums import GeometryType
if TYPE_CHECKING: # pragma: no cover
from tiatoolbox.typing import (
CallablePredicate,
CallableSelect,
Geometry,
Predicate,
Properties,
QueryGeometry,
Select,
)
sqlite3.enable_callback_tracebacks(True) # noqa: FBT003
WKB_POINT_STRUCT = struct.Struct("<BIdd")
# Only Python 3.10+ supports using slots for dataclasses
# https://docs.python.org/3/library/dataclasses.html#dataclasses.dataclass
# therefore we use the following workaround to only use them when available.
# Using slots gives a performance boost at object creation time.
USE_SLOTS = {"slots": True} if sys.version_info >= (3, 10) else {} # pragma: no cover
[docs]
@dataclass(frozen=True, init=False, eq=True, **USE_SLOTS)
class Annotation:
"""An annotation: a geometry and associated properties.
Attributes:
geometry (Geometry):
The geometry of the annotation as a Shapely object.
properties (dict):
The properties of the annotation.
wkb (bytes):
The WKB representation of the geometry.
"""
_geometry: Geometry | None = field(default=None, hash=True)
properties: Properties = field(default_factory=dict, hash=True)
_wkb: bytes | None = field(default=None, hash=False)
@property
def geometry(self: Annotation) -> Geometry:
"""Return the shapely geometry of the annotation."""
if self._geometry is None:
# Lazy creation of Shapely object when first requested. This
# is memoized under _geometry. object.__setattr__ must be
# used because the class is frozen and will disallow normal
# assignment.
object.__setattr__(self, "_geometry", shapely_wkb.loads(self._wkb))
# Return memoized geometry
return self._geometry
@property
def wkb(self: Annotation) -> bytes:
"""Return the WKB representation of the annotation."""
if self._wkb is None:
object.__setattr__(self, "_wkb", self.geometry.wkb)
return self._wkb
@property
def geometry_type(self: Annotation) -> GeometryType:
"""Return the geometry type of the annotation."""
if self._geometry:
return GeometryType(self.geometry.type)
return GeometryType(
int.from_bytes(
self._wkb[1:4],
byteorder="big" if self._wkb[0] == "b\0" else "little",
),
)
@property
def coords( # noqa: PLR0911 - 7 > 6 returns
self: Annotation,
) -> np.array | list[np.array] | list[list[np.array]]:
"""The annotation geometry as a flat array of 2D coordinates.
Returns a numpy array of coordinates for point and line string.
For polygons, returns a list of numpy arrays, one for each ring.
For multi-geometries, returns a list with one element for each
geometry.
Returns:
np.array or list:
The coordinates of the annotation geometry.
"""
if self._geometry is None:
return Annotation.decode_wkb(self._wkb, self.geometry_type.value)
geom_type = self.geometry_type
if geom_type == GeometryType.POINT:
return np.array(self.geometry.coords)
if geom_type == GeometryType.LINE_STRING:
return np.array(self.geometry.coords)
if geom_type == GeometryType.POLYGON:
return [np.array(ring.coords) for ring in shapely.get_rings(self.geometry)]
if geom_type == GeometryType.MULTI_POINT:
return [np.array(part.coords) for part in self.geometry.geoms]
if geom_type == GeometryType.MULTI_LINE_STRING:
return [np.array(part.coords) for part in self.geometry.geoms]
if geom_type == GeometryType.MULTI_POLYGON:
return [
[np.array(ring.coords) for ring in shapely.get_rings(poly)]
for poly in self.geometry.geoms
]
msg = f"Unknown geometry type: {self.geometry_type}"
raise ValueError(msg)
[docs]
def to_feature(self: Annotation) -> dict:
"""Return a feature representation of this annotation.
A feature representation is a Python dictionary with the
same schema as a geoJSON feature.
Returns:
dict:
A feature representation of this annotation.
"""
return {
"type": "Feature",
"geometry": geometry2feature(self.geometry),
"properties": self.properties,
}
[docs]
def to_geojson(self: Annotation) -> str:
"""Return a GeoJSON string representation of this annotation.
Returns:
str:
A GeoJSON representation of this annotation.
"""
return json.dumps(self.to_feature())
[docs]
def to_wkb(self: Annotation) -> bytes:
"""Returns the geometry as Well-Known Binary (WKB).
Returns:
Annotation:
The annotation as a WKB geometry.
"""
return copy.copy(self.wkb)
[docs]
def to_wkt(self: Annotation) -> str:
"""Returns the geometry as Well-Know Text (WKT).
Returns:
Annotation:
The annotation as a WKT geometry.
"""
return self.geometry.wkt
def __init__(
self: Annotation,
geometry: Geometry | int | str | None = None,
properties: Properties | None = None,
wkb: bytes | None = None,
) -> None:
"""Create a new annotation.
Must be initialized with either a Shapely geometry object or WKB
bytes.
Args:
geometry (Geometry):
The geometry of the annotation.
properties (dict):
The properties of the annotation. Optional, defaults to
{}.
wkb (bytes):
The WKB representation of a geometry. Optional.
"""
if wkb is not None and geometry is not None:
msg = "Cannot init with both geometry and wkb."
raise ValueError(msg)
if wkb is None and geometry is None:
msg = "Either geometry or wkb must be given."
raise ValueError(msg)
if wkb is not None:
object.__setattr__(self, "_wkb", wkb)
object.__setattr__(self, "_geometry", None)
else:
object.__setattr__(self, "_wkb", None)
object.__setattr__(self, "_geometry", geometry)
object.__setattr__(self, "properties", properties or {})
def __repr__(self: Annotation) -> str:
"""Return a string representation of the object."""
return f"Annotation({self.geometry}, {self.properties})"
def __hash__(self: Annotation) -> int:
"""Compute the hash value of the object.
Returns:
int:
The hash value.
"""
return hash((self.geometry, json.dumps(self.properties, sort_keys=True)))
def __eq__(self: Annotation, other: object) -> bool:
"""Compare this annotation to another.
Args:
other (object):
The object to compare to.
Returns:
bool:
True if the objects are equal, False otherwise.
"""
if not isinstance(other, Annotation):
return False
return self.geometry == other.geometry and self.properties == other.properties
[docs]
@staticmethod
def decode_wkb(
wkb: bytes,
geom_type: int,
) -> np.ndarray | list[np.ndarray] | list[list[np.ndarray]]:
r"""Decode WKB to a NumPy array of flat (alternating x, y) coordinates.
Polygons return a list of NumPy arrays (one array per ring) and
multi-part geometries return a list with one element per child
geometry e.g. a list of arrays for Multi-Line and a list of
lists of arrays for multi-polyon.
Args:
wkb (bytes):
The WKB representation of a geometry.
geom_type (int):
The type of geometry to decode. Where 1 = point, 2 =
line, 3 = polygon, 4 = multi-point, 5 = multi-line, 6 =
multi-polygon.
Examples:
>>> from tiatoolbox.annotation.storage import AnnotationStore
>>> # Point(1, 2).wkb
>>> wkb = (
... b"\x01\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00"
... b"\xf0?\x00\x00\x00\x00\x00\x00\x00@"
... )
>>> AnnotationStore.decode_wkb(wkb, 1)
array([0., 0.])
>>> from tiatoolbox.annotation.storage import AnnotationStore
>>> # Polygon([[0, 0], [1, 1], [1, 0]]).wkb
>>> wkb = (
... b"\x01\x03\x00\x00\x00\x01\x00\x00\x00\x04\x00\x00\x00"
... b"\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00"
... b"\x00\x00\x00\x00\x00\x00\x00\x00\x00\xf0?\x00\x00\x00"
... b"\x00\x00\x00\xf0?\x00\x00\x00\x00\x00\x00\xf0?\x00"
... b"\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00"
... b"\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00"
... )
>>> AnnotationStore.decode_wkb(wkb, 3)
array([[0., 0.],
[1., 1.],
[1., 0.],
[0., 0.]])
Raises:
ValueError:
If the geometry type is not supported.
Returns:
np.ndarray or list:
An array of coordinates, a list of coordinates, or a list
of lists of coordinates.
"""
def decode_polygon(offset: int = 0) -> tuple[list[np.ndarray], int]:
"""Decode a polygon from WKB.
Args:
offset (int, optional):
The starting offset in the WKB representation.
Defaults to 0.
Returns:
Tuple[np.ndarray, int]: A tuple containing the decoded
polygon rings as numpy arrays and the new offset in the
WKB representation.
"""
offset += 5 # byte order and geom type at start of each polygon
n_rings = np.frombuffer(wkb, np.int32, 1, offset)[0]
offset += 4
rings = []
for _ in range(n_rings):
n_points = np.frombuffer(wkb, np.int32, 1, offset)[0]
offset += 4
rings.append(
np.frombuffer(wkb, np.double, n_points * 2, offset).reshape(-1, 2),
)
offset += n_points * 16
return rings, offset
# Using magic numbers instead of GeometryType enums for future
# numba compilation (numba doesn't like enums).
if geom_type == 1:
# Point
return np.frombuffer(wkb, np.double, -1, 5).reshape(1, 2)
if geom_type == 2: # noqa: PLR2004 - Intentional magic number
# Line
return np.frombuffer(wkb, np.double, -1, 9).reshape(-1, 2)
if geom_type == 3: # noqa: PLR2004 - Intentional magic number
# Polygon
return decode_polygon()[0]
if geom_type == 4: # noqa: PLR2004 - Intentional magic number
# Multi-point
n_points = np.frombuffer(wkb, np.int32, 1, 5)[0]
return [
np.frombuffer(wkb, np.double, 2, 14 + i * 21).reshape(1, 2)
# each point is 21 bytes
for i in range(n_points)
]
if geom_type == 5: # noqa: PLR2004 - Intentional magic number
# Multi-line
n_lines = np.frombuffer(wkb, np.int32, 1, 5)[0]
lines = []
offset = 9
for _ in range(n_lines):
offset += 5
n_points = np.frombuffer(wkb, np.int32, n_lines, offset)[0]
offset += 4
lines.append(
np.frombuffer(wkb, np.double, n_points * 2, offset).reshape(-1, 2),
)
offset += n_points * 16
return lines
if geom_type == 6: # noqa: PLR2004 - Intentional magic number
# Multi-polygon
n_polygons = np.frombuffer(wkb, np.int32, 1, 5)[0]
polygons = []
offset = 9
for _ in range(n_polygons):
rings, offset = decode_polygon(offset)
polygons.append(rings)
return polygons
msg = f"Unknown geometry type: {geom_type}"
raise ValueError(msg)
[docs]
class AnnotationStore(ABC, MutableMapping):
"""Annotation store abstract base class."""
def __new__(cls: ABC, *args: str, **kwargs: int) -> ABC: # noqa: ARG003
"""Return an instance of a subclass of AnnotationStore."""
if cls is AnnotationStore:
msg = (
"AnnotationStore is an abstract class and cannot be instantiated."
"Use a subclass such as DictionaryStore or SQLiteStore instead."
)
raise TypeError(
msg,
)
return super().__new__(cls)
@staticmethod
def _is_right_angle(a: list[float], b: list[float], c: list[float]) -> bool:
"""Return True if three points make a right angle.
Used for optimising queries.
This function will have positional only arguments when support
for Python 3.7 is dropped.
Args:
a (Sequence[Number]):
First coordinate.
b (Sequence[Number]):
Second coordinate.
c (Sequence[Number]):
Third coordinate.
"""
return np.dot(np.subtract(a, b), np.subtract(b, c)) == 0
def _is_rectangle(
self: ABC,
a: list[float],
b: list[float],
c: list[float],
d: list[float],
*args: str,
) -> bool:
"""Determine if a set of coordinates form a rectangle.
Used for optimising queries. If more than five points are given,
or if the optional fifth point is not equal to `a` then this
returns False.
Args:
a (Sequence[Number]):
First coordinate.
b (Sequence[Number]):
Second coordinate.
c (Sequence[Number])::
Third coordinate.
d (Sequence[Number]):
Fourth coordinate.
args (list):
Non-Keyword arguments.
Returns:
True if the coordinates form a rectangle, False otherwise.
"""
# Only allow one extra coordinate for looping back to the first point
if (len(args) == 1 and not np.array_equal(args[:1], [a])) or len(args) > 1:
return False
# Check that all angles are right angles
return all(
self._is_right_angle(*xyz) for xyz in ((a, b, c), (b, c, d), (c, d, a))
)
@staticmethod
def _connection_to_path(connection: str | Path | IO) -> Path:
"""Normalise a connection object to a Path.
Here we refer to a 'connection' as anything which references a
file e.g. a string, a pathlibPath, or a file-like object (IO).
Args:
connection (Union[str, Path, IO]):
The connection object to normalise.
Returns:
Path:
The normalised path.
"""
if not isinstance(
connection,
(
str,
Path,
io.IOBase,
io.TextIOBase,
tempfile._TemporaryFileWrapper, # skipcq: PYL-W0212 # noqa: SLF001
),
):
msg = (
f"Connection must be a string, Path, or an IO object, "
f"not {type(connection)}"
)
raise TypeError(
msg,
)
if isinstance(
connection,
(
io.IOBase,
io.TextIOBase,
tempfile._TemporaryFileWrapper, # skipcq: PYL-W0212 # noqa: SLF001
),
):
connection = connection.name
return Path(connection)
@staticmethod
def _validate_equal_lengths(*args: str) -> None:
"""Validate that all given args are either None or have the same length."""
lengths = [len(v) for v in args if v is not None]
if lengths and any(length != lengths[0] for length in lengths):
msg = "All arguments must be None or of equal length."
raise ValueError(msg)
@staticmethod
def _geometry_predicate(name: str, a: Geometry, b: Geometry) -> Callable:
"""Apply a binary geometry predicate.
For more information on geometric predicates see the `Shapely
documentation <https://shapely.readthedocs.io/en/stable/
manual.html#binary-predicates>`_.
Args:
name(str):
Name of the predicate to apply.
a(Geometry):
The first geometry.
b(Geometry):
The second geometry.
Returns:
bool:
True if the geometry predicate holds.
"""
return getattr(a, name)(b)
# All valid shapely binary predicates
_geometry_predicate_names: ClassVar[list[str]] = [
"equals",
"contains",
"covers",
"covered_by",
"crosses",
"disjoint",
"intersects",
"overlaps",
"touches",
"within",
# Special non-shapely case, bounding-boxes intersect.
"bbox_intersects",
# Special non-shapely case, query centroid within k of
# annotation bounds center.
"centers_within_k",
]
[docs]
@classmethod
@abstractmethod
def open(cls: type[AnnotationStore], fp: Path | str | IO) -> ABC:
"""Load a store object from a path or file-like object.
Args:
fp(Path or str or IO): The file path or file handle.
Returns:
AnnotationStoreABC:
An instance of an annotation store.
"""
[docs]
@staticmethod
def serialise_geometry(geometry: Geometry) -> str | bytes:
"""Serialise a geometry to a string or bytes.
This defaults to well-known text (WKT) but may be overridden to
any other format which a Shapely geometry could be serialised to
e.g. well-known binary (WKB) or geoJSON etc.
Args:
geometry(Geometry):
The Shapely geometry to be serialised.
Returns:
bytes or str: The serialised geometry.
"""
return geometry.wkt
[docs]
@staticmethod
@lru_cache(32)
def deserialize_geometry(data: str | bytes) -> Geometry:
"""Deserialize a geometry from a string or bytes.
This default implementation will deserialize bytes as well-known
binary (WKB) and strings as well-known text (WKT). This can be
overridden to deserialize other formats such as geoJSON etc.
Args:
data(bytes or str):
The serialised representation of a Shapely geometry.
Returns:
Geometry: The deserialized Shapely geometry.
"""
return (
shapely_wkt.loads(data)
if isinstance(data, str)
else shapely_wkb.loads(data)
)
[docs]
@abstractmethod
def commit(self: AnnotationStore) -> None:
"""Commit any in-memory changes to disk."""
[docs]
@abstractmethod
def dump(self: AnnotationStore, fp: Path | str | IO) -> None:
"""Serialise a copy of the whole store to a file-like object.
Args:
fp(Path or str or IO):
A file path or file handle object for output to disk.
"""
[docs]
@abstractmethod
def dumps(self: AnnotationStore) -> str | bytes:
"""Serialise and return a copy of store as a string or bytes.
Returns:
str or bytes:
The serialised store.
"""
[docs]
def append(
self: AnnotationStore,
annotation: Annotation,
key: str | None = None,
) -> str:
"""Insert a new annotation, returning the key.
Args:
annotation (Annotation):
The shapely annotation to insert.
key (str):
Optional. The unique key used to identify the annotation in the
store. If not given a new UUID4 will be generated and returned
instead.
Returns:
str:
The unique key of the newly inserted annotation.
"""
if not isinstance(annotation.geometry, (Polygon, Point, LineString)):
msg = "Invalid geometry type. Must be one of Point, LineString, Polygon."
raise TypeError(msg)
keys = key if key is None else [key]
return self.append_many([annotation], keys)[0]
[docs]
def append_many(
self: AnnotationStore,
annotations: Iterable[Annotation],
keys: Iterable[str] | None = None,
) -> list[str]:
"""Bulk append of annotations.
This may be more performant than repeated calls to `append`.
Args:
annotations (iter(Annotation)):
An iterable of annotations.
keys (iter(str)):
An iterable of unique keys associated with each geometry being
inserted. If None, a new UUID4 is generated for each geometry.
Returns:
list(str):
A list of unique keys for the inserted geometries.
"""
annotations = list(annotations)
keys = list(keys) if keys else None
self._validate_equal_lengths(keys, annotations)
result = []
if keys:
result.extend(
self.append(annotation, key)
for key, annotation in zip(keys, annotations)
)
return result
result.extend(self.append(annotation) for annotation in annotations)
return result
[docs]
def patch(
self: AnnotationStore,
key: str,
geometry: Geometry | None = None,
properties: dict[str, object] | None = None,
) -> None:
"""Patch an annotation at given key.
Partial update of an annotation. Providing only a geometry will update
the geometry and leave properties unchanged. Providing a properties
dictionary applies a patch operation to the properties. Only updating
the properties which are given and leaving the rest unchanged. To
completely replace an annotation use `__setitem__`.
Args:
key(str):
The key of the annotation to update.
geometry(Geometry):
The new geometry. If None, the geometry is not updated.
properties(dict):
A dictionary of properties to patch and their new values.
If None, the existing properties are not altered.
"""
if key not in self:
self.append(Annotation(geometry, properties), key)
return
geometry = geometry if geometry is None else [geometry]
properties = properties if properties is None else [properties]
self.patch_many([key], geometry, properties)
[docs]
def patch_many(
self: AnnotationStore,
keys: Iterable[int],
geometries: Iterable[Geometry] | None = None,
properties_iter: Iterable[Properties] | None = None,
) -> None:
"""Bulk patch of annotations.
This may be more efficient than calling `patch` repeatedly
in a loop.
Args:
geometries (iter(Geometry)):
An iterable of geometries to update.
properties_iter (iter(dict)):
An iterable of properties to update.
keys (iter(str)):
An iterable of keys for each annotation to be updated.
"""
# Validate inputs
if not any([geometries, properties_iter]):
msg = "At least one of geometries or properties_iter must be given"
raise ValueError(
msg,
)
keys = list(keys)
geometries = list(geometries) if geometries else None
properties_iter = list(properties_iter) if properties_iter else None
self._validate_equal_lengths(keys, geometries, properties_iter)
properties_iter = properties_iter or ({} for _ in keys) # pragma: no branch
geometries = geometries or (None for _ in keys) # pragma: no branch
# Update the store
for key, geometry, properties in zip(keys, geometries, properties_iter):
properties_ = copy.deepcopy(properties)
self.patch(key, geometry, properties_)
[docs]
def remove(self: AnnotationStore, key: str) -> None:
"""Remove annotation from the store with its unique key.
Args:
key (str):
The key of the annotation to be removed.
"""
self.remove_many([key])
[docs]
def remove_many(self: AnnotationStore, keys: Iterable[str]) -> None:
"""Bulk removal of annotations by keys.
Args:
keys (iter(str)):
An iterable of keys for the annotation to be removed.
"""
for key in keys:
self.remove(key)
[docs]
def setdefault(
self: AnnotationStore,
key: str,
default: Annotation | None = None,
) -> Annotation:
"""Return the value of the annotation with the given key.
If the key does not exist, insert the default value and return
it.
Args:
key (str):
The key of the annotation to be fetched.
default (Annotation):
The value to return if the key is not found.
Returns:
Annotation:
The annotation or default if the key is not found.
"""
if not isinstance(default, Annotation):
msg = "default value must be an Annotation instance."
raise TypeError(msg)
return super().setdefault(key, default)
def __delitem__(self: AnnotationStore, key: str) -> None:
"""Delete an annotation by key.
An alias of `remove`.
Args:
key (str):
The key of the annotation to be removed.
"""
self.remove(key)
[docs]
def keys(self: AnnotationStore) -> Iterable[str]:
"""Return an iterable (usually generator) of all keys in the store.
Returns:
Iterable[str]:
An iterable of keys.
"""
for key, _ in self.items(): # noqa: PERF102
yield key
[docs]
def values(self: AnnotationStore) -> Iterable[Annotation]:
"""Return an iterable of all annotation in the store.
Returns:
Iterable[Annotation]:
An iterable of annotations.
"""
for _, annotation in self.items(): # noqa: PERF102
yield annotation
def __iter__(self: AnnotationStore) -> Iterator[str]:
"""Return an iterable of keys in the store.
An alias of `keys`.
Returns:
Iterator[str]:
An iterable of keys.
"""
yield from self.keys()
@staticmethod
def _eval_where(
predicate: Predicate | None,
properties: dict[str, object],
) -> bool:
"""Evaluate properties predicate against properties.
Args:
predicate (str or bytes or Callable):
The predicate to evaluate on properties. The predicate may be a
string, pickled bytes, or a Callable (e.g. a function).
properties (dict):
A dictionary of JSON serializable
properties on which to evaluate the predicate.
Returns:
bool:
Return True if the predicate holds.
"""
if predicate is None:
return True
if isinstance(predicate, str):
return bool(
eval( # skipcq: PYL-W0123, # noqa: PGH001, S307
predicate,
PY_GLOBALS,
{"props": properties},
),
)
if isinstance(predicate, bytes):
predicate = pickle.loads(predicate) # skipcq: BAN-B301 # noqa: S301
return bool(predicate(properties))
[docs]
def query(
self: AnnotationStore,
geometry: QueryGeometry | None = None,
where: Predicate | None = None,
geometry_predicate: str = "intersects",
min_area: float | None = None,
distance: float = 0,
) -> dict[str, Annotation]:
"""Query the store for annotations.
Args:
geometry (Geometry or Iterable):
Geometry to use when querying. This can be a bounds
(iterable of length 4) or a Shapely geometry (e.g.
Polygon).
where (str or bytes or Callable):
A statement which should evaluate to a boolean value.
Only annotations for which this predicate is true will
be returned. Defaults to None (assume always true). This
may be a string, Callable, or pickled function as bytes.
Callables are called to filter each result returned
from the annotation store backend in python before being
returned to the user. A pickle object is, where
possible, hooked into the backend as a user defined
function to filter results during the backend query.
Strings are expected to be in a domain specific language
and are converted to SQL on a best-effort basis. For
supported operators of the DSL see
:mod:`tiatoolbox.annotation.dsl`. E.g. a simple python
expression `props["class"] == 42` will be converted to a
valid SQLite predicate when using `SQLiteStore` and
inserted into the SQL query. This should be faster than
filtering in python after or during the query.
Additionally, the same string can be used across
different backends (e.g. the previous example predicate
string is valid for both `DictionaryStore `and a
`SQliteStore`). On the other hand it has many more
limitations. It is important to note that untrusted user
input should never be accepted to this argument as
arbitrary code can be run via pickle or the parsing of
the string statement.
geometry_predicate (str):
A string defining which binary geometry predicate to
use when comparing the query geometry and a geometry in
the store. Only annotations for which this binary
predicate is true will be returned. Defaults to
"intersects". For more information see the `shapely
documentation on binary predicates <https://shapely.
readthedocs.io/en/stable/manual.html#binary-predicates>`_.
min_area (float):
Minimum area of the annotation geometry. Only
annotations with an area greater than or equal to this
value will be returned. Defaults to None (no min).
distance (float):
Distance used when performing a distance based query.
E.g. "centers_within_k" geometry predicate.
Returns:
list:
A list of Annotation objects.
"""
if all(x is None for x in (geometry, where)):
msg = "At least one of geometry or where must be set."
raise ValueError(msg)
if geometry_predicate not in self._geometry_predicate_names:
msg = (
"Invalid geometry predicate. Allowed values are: "
f"{', '.join(self._geometry_predicate_names)}."
)
raise ValueError(
msg,
)
query_geometry = geometry
if isinstance(query_geometry, Iterable):
query_geometry = Polygon.from_bounds(*query_geometry)
if geometry_predicate == "centers_within_k":
query_point = Polygon.from_bounds(*query_geometry.bounds).centroid
def bbox_intersects(
annotation_geometry: Geometry,
query_geometry: Geometry,
) -> bool:
"""True if bounding box of the annotation intersects the query geometry."""
return Polygon.from_bounds(*query_geometry.bounds).intersects(
Polygon.from_bounds(*annotation_geometry.bounds),
)
def centers_within_k(
annotation_geometry: Geometry,
query_point: Point,
distance: float,
) -> bool:
"""True if centre of annotation within k of query geometry center.
Here the "center" is the centroid of the bounds.
"""
ann_centre = Polygon.from_bounds(*annotation_geometry.bounds).centroid
return query_point.dwithin(ann_centre, distance)
def filter_function(annotation: Annotation) -> bool:
"""Filter function for querying annotations.
Args:
annotation (Annotation):
The annotation to filter.
Returns:
bool:
True if the annotation should be included in the
query result.
"""
if min_area is not None and annotation.geometry.area < min_area:
return False
return ( # Geometry is None or the geometry predicate matches
query_geometry is None
or any(
[
(
geometry_predicate == "bbox_intersects"
and bbox_intersects(annotation.geometry, query_geometry)
),
(
geometry_predicate == "centers_within_k"
and centers_within_k(
annotation.geometry,
query_point,
distance,
)
),
(
geometry_predicate
not in ("bbox_intersects", "centers_within_k")
and self._geometry_predicate(
geometry_predicate,
query_geometry,
annotation.geometry,
)
),
],
)
) and self._eval_where(where, annotation.properties)
return {
key: annotation
for key, annotation in self.items()
if filter_function(annotation)
}
[docs]
def iquery(
self: AnnotationStore,
geometry: QueryGeometry,
where: Predicate | None = None,
geometry_predicate: str = "intersects",
) -> list[int]:
"""Query the store for annotation keys.
Acts the same as `AnnotationStore.query` except returns keys
instead of annotations.
Args:
geometry (Geometry or Iterable):
Geometry to use when querying. This can be a bounds
(iterable of length 4) or a Shapely geometry (e.g.
Polygon).
where (str or bytes or Callable):
A statement which should evaluate to a boolean value.
Only annotations for which this predicate is true will
be returned. Defaults to None (assume always true). This
may be a string, Callable, or pickled function as bytes.
Callables are called to filter each result returned
from the annotation store backend in python before being
returned to the user. A pickle object is, where
possible, hooked into the backend as a user defined
function to filter results during the backend query.
Strings are expected to be in a domain specific language
and are converted to SQL on a best-effort basis. For
supported operators of the DSL see
:mod:`tiatoolbox.annotation.dsl`. E.g. a simple python
expression `props["class"] == 42` will be converted to a
valid SQLite predicate when using `SQLiteStore` and
inserted into the SQL query. This should be faster than
filtering in python after or during the query.
Additionally, the same string can be used across
different backends (e.g. the previous example predicate
string is valid for both `DictionaryStore `and a
`SQliteStore`). On the other hand it has many more
limitations. It is important to note that untrusted user
input should never be accepted to this argument as
arbitrary code can be run via pickle or the parsing of
the string statement.
geometry_predicate:
A string which define which binary geometry predicate to
use when comparing the query geometry and a geometry in
the store. Only annotations for which this binary
predicate is true will be returned. Defaults to
"intersects". For more information see the `shapely
documentation on binary predicates <https://shapely.
readthedocs.io/en/stable/manual.html#binary-predicates>`_.
Returns:
list:
A list of keys for each Annotation.
"""
if geometry_predicate not in self._geometry_predicate_names:
msg = (
"Invalid geometry predicate. Allowed values are: "
f"{', '.join(self._geometry_predicate_names)}."
)
raise ValueError(
msg,
)
query_geometry = geometry
if isinstance(query_geometry, Iterable):
query_geometry = Polygon.from_bounds(*query_geometry)
return [
key
for key, annotation in self.items()
if (
self._geometry_predicate(
geometry_predicate,
query_geometry,
annotation.geometry,
)
and self._eval_where(where, annotation.properties)
)
]
[docs]
def bquery(
self: AnnotationStore,
geometry: QueryGeometry | None = None,
where: Predicate | None = None,
) -> dict[str, tuple[float, float, float, float]]:
"""Query the store for annotation bounding boxes.
Acts similarly to `AnnotationStore.query` except it checks for
intersection between stored and query geometry bounding boxes.
This may be faster than a regular query in some cases, e.g. for
SQliteStore with a large number of annotations.
Note that this method only checks for bounding box intersection
and therefore may give a different result to using
`AnnotationStore.query` with a box polygon and the "intersects"
geometry predicate. Also note that geometry predicates are not
supported for this method.
Args:
geometry (Geometry or Iterable):
Geometry to use when querying. This can be a bounds
(iterable of length 4) or a Shapely geometry (e.g.
Polygon). If a geometry is provided, the bounds of the
geometry will be used for the query. Full geometry
intersection is not used for the query method.
where (str or bytes or Callable):
A statement which should evaluate to a boolean value.
Only annotations for which this predicate is true will
be returned. Defaults to None (assume always true). This
may be a string, Callable, or pickled function as bytes.
Callables are called to filter each result returned
from the annotation store backend in python before being
returned to the user. A pickle object is, where
possible, hooked into the backend as a user defined
function to filter results during the backend query.
Strings are expected to be in a domain specific language
and are converted to SQL on a best-effort basis. For
supported operators of the DSL see
:mod:`tiatoolbox.annotation.dsl`. E.g. a simple python
expression `props["class"] == 42` will be converted to a
valid SQLite predicate when using `SQLiteStore` and
inserted into the SQL query. This should be faster than
filtering in python after or during the query.
Additionally, the same string can be used across
different backends (e.g. the previous example predicate
string is valid for both `DictionaryStore` and a
`SQliteStore`). On the other hand it has many more
limitations. It is important to note that untrusted user
input should never be accepted to this argument as
arbitrary code can be run via pickle or the parsing of
the string statement.
Returns:
list:
A list of bounding boxes for each Annotation.
Example:
>>> from tiatoolbox.annotation.storage import DictionaryStore
>>> from shapely.geometry import Polygon
>>> store = DictionaryStore()
>>> store.append(
... Annotation(
... geometry=Polygon.from_bounds(0, 0, 1, 1),
... properties={"class": 42},
... ),
... key="foo",
... )
>>> store.bquery(where="props['class'] == 42")
{'foo': (0.0, 0.0, 1.0, 1.0)}
"""
query_geometry = geometry
if isinstance(query_geometry, Iterable):
query_geometry = Polygon.from_bounds(*query_geometry)
return {
key: annotation.geometry.bounds
for key, annotation in self.items()
if (
query_geometry is None
or Polygon.from_bounds(*annotation.geometry.bounds).intersects(
Polygon.from_bounds(*query_geometry.bounds),
)
and self._eval_where(where, annotation.properties)
)
}
[docs]
def pquery(
self: AnnotationStore,
select: Select,
geometry: QueryGeometry | None = None,
where: Predicate | None = None,
*,
unique: bool = True,
squeeze: bool = True,
) -> dict[str, object] | set[object]:
"""Query the store for annotation properties.
Acts similarly to `AnnotationStore.query` but returns only the
value defined by `select`.
Args:
select (str or bytes or Callable):
A statement defining the value to look up from the
annotation properties. If `select = "*"`, all properties
are returned for each annotation (`unique` must be
False).
geometry (Geometry or Iterable):
Geometry to use when querying. This can be a bounds
(iterable of length 4) or a Shapely geometry (e.g.
Polygon). If a geometry is provided, the bounds of the
geometry will be used for the query. Full geometry
intersection is not used for the query method.
where (str or bytes or Callable):
A statement which should evaluate to a boolean value.
Only annotations for which this predicate is true will
be returned. Defaults to None (assume always true). This
may be a string, Callable, or pickled function as bytes.
Callables are called to filter each result returned
from the annotation store backend in python before being
returned to the user. A pickle object is, where
possible, hooked into the backend as a user defined
function to filter results during the backend query.
Strings are expected to be in a domain specific language
and are converted to SQL on a best-effort basis. For
supported operators of the DSL see
:mod:`tiatoolbox.annotation.dsl`. E.g. a simple python
expression `props["class"] == 42` will be converted to a
valid SQLite predicate when using `SQLiteStore` and
inserted into the SQL query. This should be faster than
filtering in python after or during the query. It is
important to note that untrusted user input should never
be accepted to this argument as arbitrary code can be
run via pickle or the parsing of the string statement.
unique (bool):
If True, only unique values for each selected property
will be returned as a list of sets. If False, all values
will be returned as a dictionary mapping keys values.
Defaults to True.
squeeze (bool):
If True, when querying for a single value with
`unique=True`, the result will be a single set instead
of a list of sets.
Examples:
>>> from tiatoolbox.annotation.storage import DictionaryStore
>>> from shapely.geometry import Point
>>> store = DictionaryStore()
>>> annotation = Annotation(
... geometry=Point(0, 0),
... properties={"class": 42},
... )
>>> store.append(annotation, "foo")
>>> store.pquery("*", unique=False)
... {'foo': {'class': 42}}
>>> from tiatoolbox.annotation.storage import DictionaryStore
>>> from shapely.geometry import Point
>>> store = DictionaryStore()
>>> annotation = Annotation(
... geometry=Point(0, 0),
... properties={"class": 42},
... )
>>> store.append(annotation, "foo")
>>> store.pquery("props['class']")
... {42}
>>> annotation = Annotation(Point(1, 1), {"class": 123})
>>> store.append(annotation, "foo")
>>> store.pquery("props['class']")
... {42, 123}
"""
if where is not None and type(select) is not type(where):
msg = "select and where must be of the same type"
raise TypeError(msg)
if not isinstance(select, (str, bytes)) and not callable(select):
msg = f"select must be str, bytes, or Callable, not {type(select)}"
raise TypeError(
msg,
)
# Are we scanning through all annotations?
is_scan = not any((geometry, where))
items = self.items() if is_scan else self.query(geometry, where).items()
def select_values(
select: Select,
annotation: Annotation,
) -> Properties | object | tuple[object, ...]:
"""Get the value(s) to return from an annotation via a select.
Args:
select (str or bytes or Callable):
A statement defining the value to look up from the
annotation properties. If `select = "*"`, all properties
are returned for each annotation (`unique` must be
False).
annotation (Annotation):
The annotation to get the value(s) from.
Raises:
ValueError:
If arguments have incompatible values.
Returns:
Union[Properties, object, Tuple[object, ...]]:
The value(s) to return from the annotation. This
will be a dictionary if unique is False. Otherwise,
it will be a list of sets. If squeeze and unique are
True in addition to there only being one set in the
results list, the result will be a single set.
""" # Q440, Q441
if select == "*" and unique:
msg = "unique=True cannot be used with select='*'"
raise ValueError(msg)
if select == "*": # Special case for all properties
return annotation.properties
if isinstance(select, str):
py_locals = {"props": annotation.properties}
return eval( # skipcq: PYL-W0123, # noqa: PGH001, S307
select,
PY_GLOBALS,
py_locals,
)
if isinstance(select, bytes):
return pickle.loads(select)( # skipcq: BAN-B301 # noqa: S301
annotation.properties,
)
return select(annotation.properties)
return self._handle_pquery_results(
select=select,
items=items,
get_values=select_values,
unique=unique,
squeeze=squeeze,
)
[docs]
def nquery(
self: AnnotationStore,
geometry: Geometry | None = None,
where: Predicate | None = None,
n_where: Predicate | None = None,
distance: float = 5.0,
geometry_predicate: str = "intersects",
mode: str = "poly-poly",
) -> dict[str, dict[str, Annotation]]:
"""Query for annotations within a distance of another annotation.
Args:
geometry (Geometry):
A geometry to use to query for the initial set of
annotations to perform a neighbourhood search around. If
None, all annotations in the store are considered.
Defaults to None.
where (str or bytes or Callable):
A statement which should evaluate to a boolean value.
Only annotations for which this predicate is true will be
returned. Defaults to None (assume always true). This may
be a string, Callable, or pickled function as bytes.
Callables are called to filter each result returned the
annotation store backend in python before being returned
to the user. A pickle object is, where possible, hooked
into the backend as a user defined function to filter
results during the backend query. Strings are expected to
be in a domain specific language and are converted to SQL
on a best-effort basis. For supported operators of the DSL
see :mod:`tiatoolbox.annotation.dsl`. E.g. a simple python
expression `props["class"] == 42` will be converted to a
valid SQLite predicate when using `SQLiteStore` and
inserted into the SQL query. This should be faster than
filtering in python after or during the query. It is
important to note that untrusted user input should never
be accepted to this argument as arbitrary code can be
run via pickle or the parsing of the string statement.
n_where (str or bytes or Callable):
Predicate to filter the nearest annotations by. Defaults
to None (assume always true). See `where` for more
details.
distance (float):
The distance to search for annotations within. Defaults to
5.0.
geometry_predicate (str):
The predicate to use when comparing geometries. Defaults
to "intersects". Other options include "within" and
"contains". Ignored if `mode` is "boxpoint-boxpoint" or
"box-box".
mode (tuple[str, str] or str):
The method to use for determining distance during the
query. Defaults to "box-box". This may significantly
change performance depending on the backend. Possible
options are:
- "poly-poly": Polygon boundary to polygon boundary.
- "boxpoint-boxpoint": Bounding box centre point to
bounding box centre point.
- "box-box": Bounding box to bounding box.
May be specified as a dash separated string or a tuple
of two strings. The first string is the mode for the
query geometry and the second string is the mode for
the nearest annotation geometry.
Returns:
Dict[str, Dict[str, Annotation]]:
A dictionary mapping annotation keys to another
dictionary which represents an annotation key and all
annotations within `distance` of it.
The `mode` argument is used to determine how to calculate the
distance between annotations. The default mode is "box-box".
The "box-box" mode uses the bounding boxes of stored annotations
and the query geometry when determining if annotations are
within the neighbourhood.
.. figure:: ../images/nquery-box-box.png
:width: 512
:alt: "box-box" mode
The "poly-poly" performs full polygon-polygon intersection with
the polygon boundary of stored annotations and the query
geometry to determine if annotations are within the
neighbourhood.
.. figure:: ../images/nquery-poly-poly.png
:width: 512
:alt: "poly-poly" mode
The "boxpoint-boxpoint" mode uses the centre point of the
bounding box of stored annotations and the query geometry when
determining if annotations are within the neighbourhood.
.. figure:: ../images/nquery-boxpoint-boxpoint.png
:width: 512
:alt: "boxpoint-boxpoint" mode
Examples:
Example bounding box query with one neighbour within a
distance of 2.0.
>>> from shapely.geometry import Point, Polygon
>>> from tiatoolbox.annotation.storage import Annotation, SQLiteStore
>>> store = SQLiteStore()
>>> annotation = Annotation(Point(0, 0), {"class": 42})
>>> store.append(annotation, "foo")
>>> neighbour = Annotation(Point(1, 1), {"class": 123})
>>> store.add(neighbour, "bar")
>>> store.nquery((-.5, -.5, .5, .5), distance=2.0)
{
"foo": {
Annotation(POINT (0 0), {'class': 42}): {
"bar": Annotation(POINT (1 1), {'class': 123}),
}
},
}
Example bounding box query with no neighbours within a
distance of 1.0.
>>> from shapely.geometry import Point
>>> from tiatoolbox.annotation.storage import Annotation, SQLiteStore
>>> store = SQLiteStore()
>>> annotation = Annotation(Point(0, 0), {"class": 42})
>>> store.add(annotation, "foo")
>>> store.nquery((-.5, -.5, .5, .5), distance=1.0)
{"foo": {Annotation(POINT (0 0), {'class': 42}): {}}}
Example of querying for TILs - lympocytes within 3 units
of tumour cells.
>>> from tiatoolbox.annotation.storage import SQLiteStore
>>> store = SQLiteStore("hovernet-pannuke-output.db")
>>> tils = store.nquery(
... where="props['class'] == 1", # Tumour cells
... n_where="props['class'] == 0", # Lymphocytes
... distance=32.0, # n_where within 32 units of where
... mode="point-point", # Use point to point distance
... )
"""
# This is a naive generic implementation which can be overridden
# by back ends which can do this more efficiently.
if not isinstance(mode, (str, tuple)):
msg = "mode must be a string or tuple of strings"
raise TypeError(msg)
if isinstance(mode, str):
mode = tuple(mode.split("-"))
if mode not in (("box", "box"), ("boxpoint", "boxpoint"), ("poly", "poly")):
msg = "mode must be one of 'box-box', 'boxpoint-boxpoint', or 'poly-poly'"
raise ValueError(
msg,
)
from_mode, _ = mode
# Initial selection of annotations to query around
selection = self.query(
geometry=geometry,
where=where,
)
# Query for others within the distance of initial selection
result = {}
for key, ann in selection.items():
geometry = ann.geometry
if from_mode == "box":
geometry_predicate = "bbox_intersects"
min_x, min_y, max_x, max_y = ann.geometry.bounds
geometry = Polygon.from_bounds(
min_x - distance,
min_y - distance,
max_x + distance,
max_y + distance,
)
elif from_mode == "boxpoint":
geometry_predicate = "centers_within_k"
elif from_mode == "poly": # pragma: no branch
geometry = ann.geometry
geometry = geometry.buffer(distance)
subquery_result = self.query(
geometry=geometry,
where=n_where,
geometry_predicate=geometry_predicate,
distance=distance,
)
if subquery_result:
result[key] = subquery_result
return result
@staticmethod
def _handle_pquery_results(
select: Select,
items: Generator[tuple[str, Properties], None, None],
get_values: Callable[
[Select, Annotation],
Properties | object | tuple[object, ...],
],
*,
unique: bool,
squeeze: bool,
) -> dict[str, object] | dict:
"""Package the results of a pquery into the right output format.
Args:
select (str or bytes or Callable):
A statement defining the value to look up from the
annotation properties. If `select = "*"`, all properties
are returned for each annotation (`unique` must be
False).
unique (bool):
If True, only unique values for each selected property
will be returned as a list of sets. If False, all values
will be returned as a dictionary mapping keys values.
Defaults to True.
squeeze (bool):
If True, when querying for a single value with
`unique=True`, the result will be a single set instead
of a list of sets.
items (Dict[str, Properties]):
A dictionary mapping annotation keys/IDs to annotation
properties.
get_values (Callable):
A function to get the values to return from an
annotation via a select.
""" # Q440, Q441
result = defaultdict(set) if unique else {}
for key, annotation in items:
values = get_values(select, annotation)
if unique:
# Wrap scalar values in a tuple
if not isinstance(values, tuple):
values = (values,)
# Add each value to the result set
for i, value in enumerate(values):
result[i].add(value)
else:
result[key] = values
if unique:
result = list(result.values())
if unique and squeeze and len(result) == 1:
result = result[0]
return result # CCR001
[docs]
def features(self: AnnotationStore) -> Generator[dict[str, object], None, None]:
"""Return annotations as a list of geoJSON features.
Returns:
list:
List of features as dictionaries.
"""
for a in self.values():
yield a.to_feature()
[docs]
def to_geodict(self: AnnotationStore) -> dict[str, object]:
"""Return annotations as a dictionary in geoJSON format.
Returns:
dict:
Dictionary of annotations in geoJSON format.
"""
return {
"type": "FeatureCollection",
"features": list(self.features()),
}
@staticmethod
def _dump_cases(
fp: IO | str | Path | None,
file_fn: Callable[[IO], None],
none_fn: Callable[[], str | bytes],
) -> str | bytes | None:
"""Helper function to handle cases for dumping.
Args:
fp:
The file path or handle to dump to.
file_fn(Callable):
The function to call when fp is a file handle.
none_fn(Callable):
The function to call when fp is None.
Returns:
str | bytes | None:
The result of dump. Depends on the provided functions.
"""
if fp is not None:
# It is a file-like object, write to it
if hasattr(fp, "write"):
return file_fn(fp)
# Turn a path into a file handle, then write to it
with Path(fp).open("w", encoding="utf-8") as file_handle:
return file_fn(file_handle)
# Return as str or bytes if no handle/path is given
return none_fn()
@staticmethod
def _load_cases(
fp: IO | str | Path,
string_fn: Callable[[str | bytes], object],
file_fn: Callable[[IO], object],
) -> object:
"""Loads cases for an input file handle or path."""
with contextlib.suppress(OSError):
if isinstance(fp, (Path, str)) and Path(fp).exists():
with Path(fp).open() as file_handle:
return file_fn(file_handle)
if isinstance(fp, (str, bytes)):
return string_fn(fp)
if hasattr(fp, "read"):
return file_fn(fp)
msg = "Invalid file handle or path."
raise OSError(msg)
[docs]
@classmethod
def from_geojson(
cls: type[AnnotationStore],
fp: IO | str,
scale_factor: tuple[float, float] = (1, 1),
origin: tuple[float, float] = (0, 0),
transform: Callable[[Annotation], Annotation] | None = None,
) -> AnnotationStore:
"""Create a new database with annotations loaded from a geoJSON file.
Args:
fp (Union[IO, str, Path]):
The file path or handle to load from.
scale_factor (Tuple[float, float]):
The scale factor in each dimension to use when loading the annotations.
All coordinates will be multiplied by this factor to allow import of
annotations saved at non-baseline resolution.
origin (Tuple[float, float]):
The x and y coordinates to use as the origin for the annotations.
transform (Callable):
A function to apply to each annotation after loading. Should take an
annotation as input and return an annotation. Defaults to None.
Intended to facilitate modifying the way annotations are loaded to
accomodate the specifics of different annotation formats.
Returns:
AnnotationStore:
A new annotation store with the annotations loaded from the file.
Example:
To load annotations from a GeoJSON exported by QuPath, with measurements
stored in a 'measurements' property as a list of name-value pairs, and
unpack those measurements into a flat dictionary of properties of
each annotation:
>>> from tiatoolbox.annotation.storage import SQLiteStore
>>> def unpack_qupath(ann: Annotation) -> Annotation:
>>> #Helper function to unpack QuPath measurements.
>>> props = ann.properties
>>> measurements = props.pop("measurements")
>>> for m in measurements:
>>> props[m["name"]] = m["value"]
>>> return ann
>>> store = SQLiteStore.from_geojson(
... "exported_file.geojson",
... transform=unpack_qupath,
... )
"""
store = cls()
if transform is None:
def transform(annotation: Annotation) -> Annotation:
"""Default import transform. Does Nothing."""
return annotation
store.add_from_geojson(
fp,
scale_factor,
origin=origin,
transform=transform,
)
return store
[docs]
def add_from_geojson(
self: AnnotationStore,
fp: IO | str,
scale_factor: tuple[float, float] = (1, 1),
origin: tuple[float, float] = (0, 0),
transform: Callable[[Annotation], Annotation] | None = None,
) -> None:
"""Add annotations from a .geojson file to an existing store.
Make the best effort to create valid shapely geometries from provided contours.
Args:
fp (Union[IO, str, Path]):
The file path or handle to load from.
scale_factor (float):
The scale factor to use when loading the annotations. All coordinates
will be multiplied by this factor to allow import of annotations saved
at non-baseline resolution.
origin (Tuple[float, float]):
The x and y coordinates to use as the origin for the annotations.
transform (Callable):
A function to apply to each annotation after loading. Should take an
annotation as input and return an annotation. Defaults to None.
Intended to facilitate modifying the way annotations are loaded to
accommodate the specifics of different annotation formats.
"""
def transform_geometry(geom: Geometry) -> Geometry:
"""Helper function to transform a geometry if needed."""
if origin != (0, 0):
# transform coords to be relative to given origin.
geom = translate(geom, -origin[0], -origin[1])
if scale_factor != (1, 1):
geom = scale(
geom,
xfact=scale_factor[0],
yfact=scale_factor[1],
origin=(0, 0, 0),
)
return geom
geojson = self._load_cases(
fp=fp,
string_fn=json.loads,
file_fn=json.load,
)
annotations = [
transform(
Annotation(
transform_geometry(
feature2geometry(feature["geometry"]),
),
feature["properties"],
),
)
for feature in geojson["features"]
]
logger.info("Adding %d annotations.", len(annotations))
self.append_many(annotations)
[docs]
def to_geojson(
self: AnnotationStore,
fp: IO | str | Path | None = None,
) -> str | None:
"""Serialise the store to geoJSON.
For more information on the geoJSON format see:
- https://geojson.org/
- https://tools.ietf.org/html/rfc7946
Args:
fp (IO):
A file-like object supporting `.read`. Defaults to None
which returns geoJSON as a string.
Returns:
Optional[str]:
None if writing to file or the geoJSON string if `fp` is
None.
"""
def write_geojson_to_file_handle(file_handle: IO) -> str / bytes / None:
"""Write the store to a GeoJson file give a handle.
This replaces the naive method which uses a lot of memory::
json.dump(self.to_geodict(), file_handle)
"""
# Write head
file_handle.write('{"type": "FeatureCollection", "features": [')
# Write each feature
for feature in self.features():
file_handle.write(json.dumps(feature)) # skipcq: PY-W0079
tell = file_handle.tell()
# Comma separate features
file_handle.write(",")
# Seek to before last comma
file_handle.seek(tell, os.SEEK_SET)
# Write tail
file_handle.write("]}")
return self._dump_cases(
fp=fp,
file_fn=write_geojson_to_file_handle,
none_fn=lambda: json.dumps(self.to_geodict()),
)
[docs]
def to_ndjson(self: AnnotationStore, fp: IO | None = None) -> str | None:
"""Serialise to New Line Delimited JSON.
Each line contains a JSON object with the following format:
.. code-block:: json
{
"key": "...",
"geometry": {
"type": "...",
"coordinates": [...]
},
"properties": {
"...": "..."
}
}
That is a geoJSON object with an additional key field.
For more information on the NDJSON format see:
- ndjson Specification: http://ndjson.org
- JSON Lines Documentation: https://jsonlines.org
- Streaming JSON: https://w.wiki/4Qan
- GeoJSON RFC: https://tools.ietf.org/html/rfc7946
- JSON RFC: https://tools.ietf.org/html/rfc7159
Args:
fp (IO): A file-like object supporting `.read`. Defaults to
None which returns geoJSON as a string.
Returns:
Optional[str]:
None if writing to file or the geoJSON string if`fp` is
None.
"""
string_lines_generator = (
json.dumps({"key": key, **annotation.to_feature()}, separators=(",", ":"))
+ "\n"
for key, annotation in self.items()
)
return self._dump_cases(
fp=fp,
file_fn=lambda fp: fp.writelines(string_lines_generator),
none_fn=lambda: "".join(string_lines_generator),
)
[docs]
@classmethod
def from_ndjson(cls: type[AnnotationStore], fp: IO | str) -> AnnotationStore:
"""Load annotations from NDJSON.
Expects each line to be a JSON object with the following format:
.. code-block:: json
{
"key": "...",
"geometry": {
"type": "...",
"coordinates": [...]
},
"properties": {
"...": "..."
}
}
That is a geoJSON object with an additional key field. If this key
field is missing, then a new UUID4 key will be generated for this
annotation.
Args:
fp (IO): A file-like object supporting `.read`.
Returns:
AnnotationStore:
The loaded annotations.
"""
store = cls()
for line in cls._load_cases(
fp=fp,
string_fn=lambda fp: fp.splitlines(),
file_fn=lambda fp: fp.readlines(),
):
dictionary = json.loads(line)
key = dictionary.get("key", uuid.uuid4().hex)
geometry = feature2geometry(dictionary["geometry"])
properties = dictionary["properties"]
store.append(Annotation(geometry, properties), key)
return store
[docs]
@classmethod
def from_dataframe(cls: type[AnnotationStore], df: pd.DataFrame) -> AnnotationStore:
"""Converts to AnnotationStore from :class:`pandas.DataFrame`."""
store = cls()
for key, row in df.iterrows():
geometry = row["geometry"]
properties = dict(row.filter(regex="^(?!geometry|key).*$"))
store.append(Annotation(geometry, properties), str(key))
return store
[docs]
def to_dataframe(self: AnnotationStore) -> pd.DataFrame:
"""Converts AnnotationStore to :class:`pandas.DataFrame`."""
features = (
{
"geometry": annotation.geometry,
"properties": annotation.properties,
"key": key,
}
for key, annotation in self.items()
)
return pd.json_normalize(features).set_index("key")
def __del__(self: AnnotationStore) -> None:
"""Implements destructor method.
This should be called when all references to the object have been deleted
i.e., when an object is garbage collected.
"""
self.close()
[docs]
def clear(self: AnnotationStore) -> None:
"""Remove all annotations from the store.
This is a naive implementation, it simply iterates over all annotations
and removes them. Faster implementations may be possible in specific
cases and may be implemented by subclasses.
"""
for key in list(self.keys()):
del self[key]
[docs]
class SQLiteStore(AnnotationStore):
"""SQLite backed annotation store.
Uses and rtree index for fast spatial queries.
Version History:
1.0.0:
Initial version.
1.0.1 (07/10/2022):
Added optional "area" column and queries sorted/filtered by area.
"""
[docs]
@classmethod
def open(cls: type[SQLiteStore], fp: Path | str) -> SQLiteStore:
"""Opens :class:`SQLiteStore` from file pointer or path."""
return SQLiteStore(fp)
def __init__(
self: SQLiteStore,
connection: Path | str | IO = ":memory:",
compression: str = "zlib",
compression_level: int = 9,
*,
auto_commit: bool = True,
) -> None:
"""Initialize :class:`SQLiteStore`."""
super().__init__()
# Check that JSON and RTree support is enabled
compile_options = self.compile_options()
if sqlite3.sqlite_version_info >= (3, 38, 0):
if not all(
["OMIT_JSON" not in compile_options, "ENABLE_RTREE" in compile_options],
):
msg = (
"RTREE sqlite3 compile option is required, and\n"
"JSON must not be disabled with OMIT_JSON compile option"
)
raise OSError(
msg,
)
elif not all(
["ENABLE_JSON1" in compile_options, "ENABLE_RTREE" in compile_options],
):
msg = "RTREE and JSON1 sqlite3 compile options are required."
raise OSError(msg)
# Check that math functions are enabled
if "ENABLE_MATH_FUNCTIONS" not in compile_options:
logger.warning(
"SQLite math functions are not enabled."
" This may cause problems with some queries."
" For example, floor division (//) will not work."
" For a full list see https://tia-toolbox.readthedocs.io/"
"en/v%s/_autosummary/tiatoolbox.annotation.dsl.html",
tiatoolbox.__version__,
)
# Set up database connection and cursor
self.connection = connection
self.path = self._connection_to_path(self.connection)
self.auto_commit = auto_commit
# Check if the path is a non-empty file
exists = (
# Use 'and' to short-circuit
self.path.is_file() and self.path.stat().st_size > 0
)
self.cons = {}
self.con.execute("BEGIN")
# Set up metadata
self.metadata = SQLiteMetadata(self.con)
if not exists:
self.metadata["version"] = "1.0.1"
self.metadata["compression"] = compression
self.metadata["compression_level"] = compression_level
# store locally as constantly fetching from db in (de)serialization is slow
self.compression = self.metadata["compression"]
self.compression_level = self.metadata["compression_level"]
if exists:
self.table_columns = self._get_table_columns()
return
# Create tables for geometry and RTree index
self.con.execute(
"""
CREATE VIRTUAL TABLE rtree USING rtree(
id, -- Integer primary key
min_x, max_x, -- 1st dimension min, max
min_y, max_y -- 2nd dimension min, max
)
""",
)
self.con.execute(
"""
CREATE TABLE annotations(
id INTEGER PRIMARY KEY, -- Integer primary key
key TEXT UNIQUE, -- Unique identifier (UUID)
objtype TEXT, -- Object type
cx FLOAT NOT NULL, -- X of centroid/representative point
cy FLOAT NOT NULL, -- Y of centroid/representative point
geometry BLOB, -- Detailed geometry
properties TEXT, -- JSON properties
area FLOAT NOT NULL -- Area (for ordering)
)
""",
)
if self.auto_commit:
self.con.commit()
self.table_columns = self._get_table_columns()
def __getattribute__(self: SQLiteStore, name: str) -> Any: # noqa: ANN401
"""If attr is con, return thread-local connection."""
if name == "con":
return self.get_connection(threading.get_ident())
return super().__getattribute__(name)
[docs]
def get_connection(self: SQLiteStore, thread_id: int) -> sqlite3.Connection:
"""Get a connection to the database."""
if thread_id not in self.cons:
con = sqlite3.connect(str(self.path), isolation_level="DEFERRED", uri=True)
# Register predicate functions as custom SQLite functions
def wkb_predicate(
name: str,
wkb_a: bytes,
b: bytes,
cx: float,
cy: float,
) -> bool:
"""Wrapper function to allow WKB as inputs to binary predicates."""
a = shapely_wkb.loads(wkb_a)
b = self._unpack_geometry(b, cx, cy)
return self._geometry_predicate(name, a, b)
def pickle_expression(pickle_bytes: bytes, properties: str) -> bool:
"""Function to load and execute pickle bytes with "properties" dict."""
fn = pickle.loads(pickle_bytes) # skipcq: BAN-B301 # noqa: S301
properties = json.loads(properties)
return fn(properties)
def get_area(wkb_bytes: bytes, cx: float, cy: float) -> float:
"""Function to get the area of a geometry."""
return self._unpack_geometry(
wkb_bytes,
cx,
cy,
).area
# Register custom functions
def register_custom_function(
name: str,
nargs: int,
fn: Callable,
*,
deterministic: bool = False,
) -> None:
"""Register a custom SQLite function.
Only Python >= 3.8 supports deterministic functions,
fallback to without this argument if not available.
Args:
name:
The name of the function.
nargs:
The number of arguments the function takes.
fn:
The function to register.
deterministic:
Whether the function is deterministic.
"""
con.create_function(
name,
nargs,
fn,
deterministic=deterministic,
)
register_custom_function(
"geometry_predicate",
5,
wkb_predicate,
deterministic=True,
)
register_custom_function(
"pickle_expression",
2,
pickle_expression,
deterministic=True,
)
register_custom_function("REGEXP", 2, py_regexp)
register_custom_function("REGEXP", 3, py_regexp)
register_custom_function("LISTSUM", 1, json_list_sum)
register_custom_function("CONTAINS", 1, json_contains)
register_custom_function("get_area", 3, get_area)
self.cons[thread_id] = con
return con
return self.cons[thread_id]
[docs]
def serialise_geometry( # skipcq: PYL-W0221
self: SQLiteStore,
geometry: Geometry,
) -> str | bytes:
"""Serialise a geometry to WKB with optional compression.
Converts shapely geometry objects to well-known binary (WKB) and
applies optional compression.
Args:
geometry(Geometry):
The Shapely geometry to be serialised.
Returns:
bytes or str:
The serialised geometry.
"""
data = geometry.wkb
if self.compression is None:
return data
if self.compression == "zlib":
return zlib.compress(data, level=self.compression_level)
msg = "Unsupported compression method."
raise ValueError(msg)
def _unpack_geometry(
self: SQLiteStore,
data: str | bytes,
cx: float,
cy: float,
) -> Geometry:
"""Return the geometry using WKB data and rtree bounds index.
For space optimisation, points are stored as centroids and all
other geometry types are stored as WKB. This function unpacks
the WKB data and uses the rtree index to find the centroid for
points where the data is null.
Args:
data (bytes or str):
The WKB/WKT data to be unpacked.
cx (int):
The X coordinate of the centroid/representative point.
cy (float):
The Y coordinate of the centroid/representative point.
Returns:
Geometry:
The Shapely geometry.
"""
return Point(cx, cy) if data is None else self.deserialize_geometry(data)
def _unpack_wkb(
self: SQLiteStore,
data: bytes,
cx: float,
cy: float,
) -> bytes:
"""Unpack WKB data."""
return (
self._decompress_data(data)
if data
else WKB_POINT_STRUCT.pack(1, GeometryType.POINT, cx, cy)
)
[docs]
def deserialize_geometry( # skipcq: PYL-W0221
self: SQLiteStore,
data: str | bytes,
) -> Geometry:
"""Deserialize a geometry from a string or bytes.
Args:
data(bytes or str):
The serialised representation of a Shapely geometry.
Returns:
Geometry:
The deserialized Shapely geometry.
"""
data = self._decompress_data(data)
if isinstance(data, str):
return shapely_wkt.loads(data)
return shapely_wkb.loads(data)
def _decompress_data(self: SQLiteStore, data: bytes) -> bytes:
"""Decompresses geometry data.
Args:
data (bytes):
The data to be decompressed.
Returns:
bytes:
The decompressed data.
Raises:
ValueError:
If the compression method is unsupported.
"""
if self.compression == "zlib":
data = zlib.decompress(data)
elif self.compression is not None:
msg = "Unsupported compression method."
raise ValueError(msg)
return data
[docs]
@staticmethod
def compile_options() -> list[str]:
"""Get the list of options that sqlite3 was compiled with.
Example:
>>> for opt in SQLiteRTreeStore.compile_options():
>>> print(opt)
COMPILER=gcc-7.5.0
ENABLE_COLUMN_METADATA
ENABLE_DBSTAT_VTAB
ENABLE_FTS3
ENABLE_FTS3_PARENTHESIS
ENABLE_FTS3_TOKENIZER
ENABLE_FTS4
ENABLE_FTS5
ENABLE_JSON1
ENABLE_LOAD_EXTENSION
ENABLE_PREUPDATE_HOOK
ENABLE_RTREE
ENABLE_SESSION
ENABLE_STMTVTAB
ENABLE_UNLOCK_NOTIFY
ENABLE_UPDATE_DELETE_LIMIT
HAVE_ISNAN
LIKE_DOESNT_MATCH_BLOBS
MAX_SCHEMA_RETRY=25
MAX_VARIABLE_NUMBER=250000
OMIT_LOOKASIDE
SECURE_DELETE
SOUNDEX
TEMP_STORE=1
THREADSAFE=1
"""
with sqlite3.connect(":memory:") as conn:
conn.enable_load_extension(True) # noqa: FBT003
options = conn.execute("pragma compile_options").fetchall()
return [opt for (opt,) in options]
[docs]
def close(self: SQLiteStore) -> None:
"""Closes :class:`SQLiteStore` from file pointer or path."""
if self.auto_commit:
self.con.commit()
self.optimize(vacuum=False, limit=1000)
for con in self.cons.values():
con.close()
self.cons = {}
def _make_token(self: SQLiteStore, annotation: Annotation, key: str | None) -> dict:
"""Create token data dict for tokenized SQL transaction."""
key = key or str(uuid.uuid4())
geometry = annotation.geometry
if geometry.geom_type == "Point":
serialised_geometry = None
else:
serialised_geometry = self.serialise_geometry(geometry)
return {
"key": key,
"geometry": serialised_geometry,
"cx": geometry.centroid.x,
"cy": geometry.centroid.y,
"min_x": geometry.bounds[0],
"min_y": geometry.bounds[1],
"max_x": geometry.bounds[2],
"max_y": geometry.bounds[3],
"geom_type": geometry.geom_type,
"properties": json.dumps(annotation.properties, separators=(",", ":")),
"area": geometry.area,
}
[docs]
def append_many(
self: SQLiteStore,
annotations: Iterable[Annotation],
keys: Iterable[str] | None = None,
) -> list[str]:
"""Appends new annotations to specified keys."""
annotations = list(annotations)
keys = list(keys) if keys else [str(uuid.uuid4()) for _ in annotations]
self._validate_equal_lengths(keys, annotations)
cur = self.con.cursor()
if self.auto_commit:
cur.execute("BEGIN")
result = []
for annotation, key in zip(annotations, keys):
self._append(key, annotation, cur)
result.append(key)
if self.auto_commit:
self.con.commit()
return result
def _append(
self: SQLiteStore,
key: str,
annotation: Annotation,
cur: sqlite3.Cursor,
) -> None:
"""Append without starting a transaction.
Args:
key(str):
The unique identifier (UUID) for the annotation.
annotation(Annotation):
The annotation to be appended.
cur(sqlite3.Cursor):
The cursor to use for the transaction.
"""
token = self._make_token(
annotation=annotation,
key=key,
)
cur.execute(
"""
INSERT INTO annotations VALUES(
NULL, :key, :geom_type,
:cx, :cy, :geometry, :properties, :area
)
""",
token,
)
row_id = cur.lastrowid
token.update({"row_id": row_id})
cur.execute(
"""
INSERT INTO rtree VALUES(
:row_id, :min_x, :max_x, :min_y, :max_y
)
""",
token,
)
@staticmethod
def _initialize_query_string_parameters(
query_geometry: Geometry | None,
query_parameters: dict[str, object],
geometry_predicate: str | None,
columns: str,
where: bytes | str,
distance: float = 0,
) -> tuple[str, dict[str, object]]:
"""Initialises the query string and parameters."""
query_string = (
"SELECT " # skipcq: BAN-B608 # noqa: S608
+ columns # skipcq: BAN-B608
+ """
FROM annotations, rtree
WHERE annotations.id == rtree.id
"""
)
# There is query geometry, add a simple rtree bounds check to
# rapidly narrow candidates down.
if query_geometry is not None:
# Add rtree index checks to the query.
# For special case of centers_within_k, Check for
# center of the annotation bounds within query geometry
# centroid + k.
if geometry_predicate == "centers_within_k":
# Use rtree index to check distance between points
query_string += (
"AND (((:min_x + :max_x)/2 - (min_x + max_x)/2)*"
"((:min_x + :max_x)/2 - (min_x + max_x)/2) + "
" ((:min_y + :max_y)/2 - (min_y + max_y)/2)*"
"((:min_y + :max_y)/2 - (min_y+ max_y)/2)) < :distance2 "
)
query_parameters["distance2"] = distance**2
# Otherwise, perform a regular bounding box intersection
else:
query_string += (
"AND max_x >= :min_x "
"AND min_x <= :max_x "
"AND max_y >= :min_y "
"AND min_y <= :max_y "
)
# Find the bounds of the geometry for the rtree index
min_x, min_y, max_x, max_y = query_geometry.bounds
# Update query parameters
query_parameters.update(
{
"min_x": min_x,
"max_x": max_x,
"min_y": min_y,
"max_y": max_y,
"geometry_predicate": geometry_predicate,
"query_geometry": query_geometry.wkb,
},
)
# The query is a full intersection check, not a simple bounds
# check only.
if geometry_predicate is not None and geometry_predicate not in (
"bbox_intersects",
"centers_within_k",
):
query_string += (
"\nAND geometry_predicate("
":geometry_predicate, :query_geometry, geometry, cx, cy"
") "
)
query_parameters["geometry_predicate"] = geometry_predicate
query_parameters["query_geometry"] = query_geometry.wkb
# Predicate is pickled function
if isinstance(where, bytes):
query_string += "\nAND pickle_expression(:where, properties)"
query_parameters["where"] = where
# Predicate is a string
if isinstance(where, str):
sql_predicate = eval( # skipcq: PYL-W0123, # noqa: PGH001, S307
where,
SQL_GLOBALS,
{},
)
query_string += f" AND {sql_predicate}"
return query_string, query_parameters
def _query( # noqa: PLR0913
self: SQLiteStore,
columns: str,
geometry: Geometry | None = None,
callable_columns: str | None = None,
geometry_predicate: str = "intersects",
where: Predicate | None = None,
min_area: float | None = None,
distance: float = 0,
*,
unique: bool = False,
no_constraints_ok: bool = False,
index_warning: bool = False,
) -> sqlite3.Cursor:
"""Common query construction logic for `query` and `iquery`.
Args:
columns(str):
The columns to select.
geometry(tuple or Geometry):
The geometry being queried against.
callable_columns(str):
The columns to select when a Callable is given to
`where`.
geometry_predicate(str):
A string which define which binary geometry predicate to
use when comparing the query geometry and a geometry in
the store. Only annotations for which this binary
predicate is true will be returned. Defaults to
"intersects". For more information see the `shapely
documentation on binary predicates <https://shapely.
readthedocs.io/en/stable/manual.html#binary-predicates>`_.
where (str or bytes or Callable):
The predicate to evaluate against candidate properties
during the query.
unique(bool):
Whether to return only unique results. Defaults to
False.
no_constraints_ok(bool):
Whether to allow the query to return results without
constraints (e.g. when the geometry or where predicate
is not provided). Defaults to False.
index_warning(bool):
Whether to warn if the query is not using an index.
Defaults to False.
min_area (float or None):
Minimum area of the annotations to be returned.
Defaults to None.
distance (float):
Distance used when performing a distance based query.
E.g. "centers_within_k" geometry predicate.
Returns:
sqlite3.Cursor:
A database cursor for the current query.
"""
if not no_constraints_ok and all(x is None for x in (geometry, where)):
msg = "At least one of `geometry` or `where` must be specified."
raise ValueError(msg)
query_geometry = geometry
if callable_columns is None:
callable_columns = columns
if geometry_predicate not in self._geometry_predicate_names:
msg = (
"Invalid geometry predicate. Allowed values are: "
f"{', '.join(self._geometry_predicate_names)}."
)
raise ValueError(
msg,
)
cur = self.con.cursor()
# Normalise query geometry and determine if it is a rectangle
if isinstance(query_geometry, Iterable):
query_geometry = Polygon.from_bounds(*query_geometry)
if isinstance(where, Callable):
columns = callable_columns
query_parameters = {}
query_string, query_parameters = self._initialize_query_string_parameters(
query_geometry,
query_parameters,
geometry_predicate,
columns,
where,
distance=distance,
)
# Add area column constraint to query if min_area is specified
if min_area is not None and "area" in self.table_columns:
query_string += f"\nAND area > {min_area}"
elif min_area is not None:
msg = (
"Cannot use `min_area` without an area column.\n"
"SQLiteStore.add_area_column() can be used to add an area column."
)
raise ValueError(
msg,
)
if unique:
query_string = query_string.replace("SELECT", "SELECT DISTINCT")
# Warn if the query is not using an index
if index_warning:
query_plan = cur.execute(
"EXPLAIN QUERY PLAN " + query_string,
query_parameters,
).fetchone()
if "USING INDEX" not in query_plan[-1]:
logger.warning(
"Query is not using an index. "
"Consider adding an index to improve performance.",
stacklevel=2,
)
# if area column exists, sort annotations by area
if "area" in self.table_columns:
query_string += "\nORDER BY area DESC"
cur.execute(query_string, query_parameters)
return cur
[docs]
def iquery(
self: SQLiteStore,
geometry: QueryGeometry | None = None,
where: Predicate | None = None,
geometry_predicate: str = "intersects",
distance: float = 0,
) -> list[str]:
"""Query the store for annotation keys.
Acts the same as `AnnotationStore.query` except returns keys
instead of annotations.
Args:
geometry (Geometry or Iterable):
Geometry to use when querying. This can be a bounds
(iterable of length 4) or a Shapely geometry (e.g.
Polygon).
where (str or bytes or Callable):
A statement which should evaluate to a boolean value.
Only annotations for which this predicate is true will
be returned. Defaults to None (assume always true). This
may be a string, Callable, or pickled function as bytes.
Callables are called to filter each result returned
from the annotation store backend in python before being
returned to the user. A pickle object is, where
possible, hooked into the backend as a user defined
function to filter results during the backend query.
Strings are expected to be in a domain specific language
and are converted to SQL on a best-effort basis. For
supported operators of the DSL see
:mod:`tiatoolbox.annotation.dsl`. E.g. a simple python
expression `props["class"] == 42` will be converted to a
valid SQLite predicate when using `SQLiteStore` and
inserted into the SQL query. This should be faster than
filtering in python after or during the query.
Additionally, the same string can be used across
different backends (e.g. the previous example predicate
string is valid for both `DictionaryStore `and a
`SQliteStore`). On the other hand it has many more
limitations. It is important to note that untrusted user
input should never be accepted to this argument as
arbitrary code can be run via pickle or the parsing of
the string statement.
geometry_predicate (str):
A string which define which binary geometry predicate to
use when comparing the query geometry and a geometry in
the store. Only annotations for which this binary
predicate is true will be returned. Defaults to
"intersects". For more information see the `shapely
documentation on binary predicates <https://shapely.
readthedocs.io/en/stable/manual.html#binary-predicates>`_.
distance (float):
Distance used when performing a distance based query.
E.g. "centers_within_k" geometry predicate.
Returns:
list:
A list of keys for each Annotation.
"""
query_geometry = geometry
cur = self._query(
"[key]",
geometry=query_geometry,
geometry_predicate=geometry_predicate,
where=where,
callable_columns="[key], properties",
distance=distance,
)
if isinstance(where, Callable):
return [
key
for key, properties in cur.fetchall()
if where(json.loads(properties))
]
return [key for (key,) in cur.fetchall()]
[docs]
def query(
self: SQLiteStore,
geometry: QueryGeometry | None = None,
where: Predicate | None = None,
geometry_predicate: str = "intersects",
min_area: float / None = None,
distance: float = 0,
) -> dict[str, Annotation]:
"""Runs Query."""
query_geometry = geometry
cur = self._query(
columns="[key], properties, cx, cy, geometry",
geometry=query_geometry,
geometry_predicate=geometry_predicate,
where=where,
min_area=min_area,
distance=distance,
)
if isinstance(where, Callable):
return {
key: Annotation(
properties=json.loads(properties),
wkb=self._unpack_wkb(blob, cx, cy),
)
for key, properties, cx, cy, blob in cur.fetchall()
if where(json.loads(properties))
}
return {
key: Annotation(
properties=json.loads(properties),
wkb=self._unpack_wkb(blob, cx, cy),
)
for key, properties, cx, cy, blob in cur.fetchall()
}
[docs]
def bquery(
self: SQLiteStore,
geometry: QueryGeometry | None = None,
where: Predicate | None = None,
) -> dict[str, tuple[float, float, float, float]]:
"""Query the store for annotation bounding boxes.
Acts similarly to `AnnotationStore.query` except it checks for
intersection between stored and query geometry bounding boxes.
This may be faster than a regular query in some cases, e.g. for
SQliteStore with a large number of annotations.
Note that this method only checks for bounding box intersection
and therefore may give a different result to using
`AnnotationStore.query` with a box polygon and the "intersects"
geometry predicate. Also note that geometry predicates are not
supported for this method.
Args:
geometry (Geometry or Iterable):
Geometry to use when querying. This can be a bounds
(iterable of length 4) or a Shapely geometry (e.g.
Polygon). If a geometry is provided, the bounds of the
geometry will be used for the query. Full geometry
intersection is not used for the query method.
where (str or bytes or Callable):
A statement which should evaluate to a boolean value.
Only annotations for which this predicate is true will
be returned. Defaults to None (assume always true). This
may be a string, Callable, or pickled function as bytes.
Callables are called to filter each result returned
from the annotation store backend in python before being
returned to the user. A pickle object is, where
possible, hooked into the backend as a user defined
function to filter results during the backend query.
Strings are expected to be in a domain specific language
and are converted to SQL on a best-effort basis. For
supported operators of the DSL see
:mod:`tiatoolbox.annotation.dsl`. E.g. a simple python
expression `props["class"] == 42` will be converted to a
valid SQLite predicate when using `SQLiteStore` and
inserted into the SQL query. This should be faster than
filtering in python after or during the query.
Additionally, the same string can be used across
different backends (e.g. the previous example predicate
string is valid for both `DictionaryStore` and a
`SQliteStore`). On the other hand it has many more
limitations. It is important to note that untrusted user
input should never be accepted to this argument as
arbitrary code can be run via pickle or the parsing of
the string statement.
Returns:
list:
A list of bounding boxes for each Annotation.
Example:
>>> from tiatoolbox.annotation.storage import SQLiteStore
>>> from shapely.geometry import Polygon
>>> store = SQLiteStore()
>>> store.append(
... Annotation(
... geometry=Polygon.from_bounds(0, 0, 1, 1),
... properties={"class": 42},
... ),
... key="foo",
... )
>>> store.bquery(where="props['class'] == 42")
{'foo': (0.0, 0.0, 1.0, 1.0)}
"""
cur = self._query(
columns="[key], min_x, min_y, max_x, max_y",
geometry=geometry,
geometry_predicate="bbox_intersects",
where=where,
callable_columns="[key], properties, min_x, min_y, max_x, max_y",
)
if isinstance(where, Callable):
return {
key: bounds
for key, properties, *bounds in cur.fetchall()
if where(json.loads(properties))
}
return {key: bounds for key, *bounds in cur.fetchall()}
@staticmethod
def _handle_pickle_callable_pquery(
select: CallableSelect,
where: CallablePredicate | None,
cur: sqlite3.Cursor,
*,
unique: bool,
) -> dict[str, set[Properties]] | dict[str, Properties]:
"""Package the results of a pquery into the right output format.
This variant is used when select and where are Callable or
pickle objects.
Args:
select (Union[str, bytes, Callable]):
A Callable to select the properties to return.
where (CallablePredicate):
A Callable predicate to filter the rows with. Maybe
None for no-op (no filtering).
cur (sqlite3.Cursor):
The cursor for the query.
unique (bool):
Whether to return only unique results.
Returns:
dict:
If unique, a dictionary of sets is returned. Otherwise,
a dictionary mapping annotation keys to JSON-like
property dictionaries is returned.
"""
def add_props_to_result(
result: defaultdict[str, set],
properties: dict[str, object],
) -> None:
"""Add the properties to the appropriate set in result.
Args:
result (DefaultDict[str, set]):
The result dictionary to add the properties to.
properties (Dict[str, object]):
The properties to add to the result.
"""
# Get the selected values
selection = select(properties)
# Wrap scalar values into a tuple
if not isinstance(selection, tuple):
selection = (selection,)
# Add the properties to the appropriate set
for i, value in enumerate(selection):
result[i].add(value)
# Load a pickled select function
if isinstance(select, bytes):
select = pickle.loads(select) # skipcq: BAN-B301 # noqa: S301
if unique:
# Create a dictionary of sets to store the unique properties
# for each property key / name.
result = defaultdict(set)
for (properties_string,) in cur.fetchall():
properties = json.loads(properties_string)
# Apply where filter and skip if False
if where and not where(properties):
continue
add_props_to_result(result, properties)
return list(result.values())
if not where:
return {
key: select(json.loads(properties))
for key, properties in cur.fetchall()
}
return {
key: select(json.loads(properties))
for key, properties in cur.fetchall()
if where(json.loads(properties))
}
@staticmethod
def _handle_str_pquery(
cur: sqlite3.Cursor,
*,
unique: bool,
star_query: bool,
) -> dict[str, set[Properties]] | dict[str, Properties]:
"""Package the results of a pquery into the right output format.
This variant is used when select and where are DSL strings.
Args:
cur (sqlite3.Cursor):
The cursor for the query.
unique (bool):
Whether to return only unique results.
star_query (bool):
True if the query is a star query, i.e. select == "*".
Returns:
dict:
If unique, a dictionary of sets is returned. Otherwise,
a dictionary mapping annotation keys to JSON-like
property dictionaries is returned.
"""
if unique:
result = defaultdict(set)
for values in cur.fetchall():
for i, value in enumerate(values):
result[i].add(value)
return list(result.values())
return {key: json.loads(x) if star_query else x for key, x in cur.fetchall()}
@staticmethod
def _kind_of_pquery(
select: Select,
where: Predicate,
) -> tuple[bool, bool, bool]:
"""Determine boolean flags for the kind of pquery this is.
If either one of `select` or `where` is a str, bytes, or
Callable, then is_callable_query, is_pickle_query, and
is_str_query respectively will be set to True.
Returns:
tuple:
A tuple of bools:
- True if select or where are Callable (functions).
- True if select or where are bytes (pickle expressions).
- True if select or where are str (SQL expressions).
"""
is_callable_query = any(isinstance(x, Callable) for x in (select, where) if x)
is_pickle_query = any(isinstance(x, bytes) for x in (select, where) if x)
is_str_query = any(isinstance(x, str) for x in (select, where) if x)
return is_callable_query, is_pickle_query, is_str_query
@staticmethod
def _validate_select_where_type(
select: Select,
where: Predicate,
) -> None:
"""Validate that select and where are valid types.
1. Check that select and where are the same type if where is given.
2. Check that select is in (str, bytes, Callable).
Raises:
TypeError:
If select and where are not the same type or not in
(str, bytes, Callable).
"""
if where is not None and type(select) is not type(where):
msg = "select and where must be of the same type"
raise TypeError(msg)
if not isinstance(select, (str, bytes)) and not callable(select):
msg = f"select must be str, bytes, or Callable, not {type(select)}"
raise TypeError(
msg,
)
[docs]
def pquery(
self: SQLiteStore,
select: Select,
geometry: QueryGeometry | None = None,
where: Predicate | None = None,
geometry_predicate: str = "intersects",
*,
unique: bool = True,
squeeze: bool = True,
) -> dict[str, object] | set[object]:
"""Query the store for annotation properties.
Acts similarly to `AnnotationStore.query` but returns only the
value defined by `select`.
Args:
select (str or bytes or Callable):
A statement defining the value to look up from the
annotation properties. If `select = "*"`, all properties
are returned for each annotation (`unique` must be
False).
geometry (Geometry or Iterable):
Geometry to use when querying. This can be a bounds
(iterable of length 4) or a Shapely geometry (e.g.
Polygon). If a geometry is provided, the bounds of the
geometry will be used for the query. Full geometry
intersection is not used for the query method.
where (str or bytes or Callable):
A statement which should evaluate to a boolean value.
Only annotations for which this predicate is true will
be returned. Defaults to None (assume always true). This
may be a string, Callable, or pickled function as bytes.
Callables are called to filter each result returned
from the annotation store backend in python before being
returned to the user. A pickle object is, where
possible, hooked into the backend as a user defined
function to filter results during the backend query.
Strings are expected to be in a domain specific language
and are converted to SQL on a best-effort basis. For
supported operators of the DSL see
:mod:`tiatoolbox.annotation.dsl`. E.g. a simple python
expression `props["class"] == 42` will be converted to a
valid SQLite predicate when using `SQLiteStore` and
inserted into the SQL query. This should be faster than
filtering in python after or during the query. It is
important to note that untrusted user input should never
be accepted to this argument as arbitrary code can be
run via pickle or the parsing of the string statement.
geometry_predicate (str):
A string defining which binary geometry predicate to
use when comparing the query geometry and a geometry in
the store. Only annotations for which this binary
predicate is true will be returned. Defaults to
"intersects". For more information see the `shapely
documentation on binary predicates <https://shapely.
readthedocs.io/en/stable/manual.html#binary-predicates>`_.
unique (bool):
If True, only unique values for each selected property
will be returned as a list of sets. If False, all values
will be returned as a dictionary mapping keys values.
Defaults to True.
squeeze (bool):
If True, when querying for a single value with
`unique=True`, the result will be a single set instead
of a list of sets.
Examples:
>>> from tiatoolbox.annotation.storage import SQLiteStore
>>> from shapely.geometry import Point
>>> store = SQLiteStore()
>>> annotation = Annotation(
... geometry=Point(0, 0),
... properties={"class": 42},
... )
>>> store.append(annotation, "foo")
>>> store.pquery("*", unique=False)
... {'foo': {'class': 42}}
>>> from tiatoolbox.annotation.storage import SQLiteStore
>>> from shapely.geometry import Point
>>> store = SQLiteStore()
>>> annotation = Annotation(
... geometry=Point(0, 0),
... properties={"class": 42},
... )
>>> store.append(annotation, "foo")
>>> store.pquery("props['class']")
... {42}
>>> annotation = Annotation(Point(1, 1), {"class": 123})
>>> store.append(annotation, "foo")
>>> store.pquery("props['class']")
... {42, 123}
"""
self._validate_select_where_type(select, where)
is_callable_query, is_pickle_query, is_str_query = self._kind_of_pquery(
select,
where,
)
is_star_query = select == "*" # Get all properties, special case
query_geometry = geometry # Rename arg
return_columns = [] # Initialise return rows list of column names
if is_star_query and unique:
msg = "unique=True cannot be used with select='*'"
raise ValueError(msg)
if not unique:
return_columns.append("[key]")
if is_str_query and not is_star_query:
select_names = eval( # skipcq: PYL-W0123, # noqa: PGH001, S307
select,
SQL_GLOBALS,
{},
)
return_columns += [str(select_names)]
if is_callable_query or is_star_query or is_pickle_query:
return_columns.append("properties")
columns = ", ".join(return_columns)
cur = self._query(
columns=columns,
geometry=query_geometry,
geometry_predicate=geometry_predicate,
where=where,
unique=unique,
no_constraints_ok=True,
index_warning=True,
)
if is_pickle_query or is_callable_query:
# Where to apply after database query
# only done for Callable where.
post_where = where if is_callable_query else None
result = self._handle_pickle_callable_pquery(
select,
post_where,
cur,
unique=unique,
)
else:
result = self._handle_str_pquery(
cur,
unique=unique,
star_query=is_star_query,
)
if unique and squeeze and len(result) == 1:
return result[0]
return result
def __len__(self: SQLiteStore) -> int:
"""Return number of annotations in the store."""
cur = self.con.cursor()
cur.execute("SELECT COUNT(*) FROM annotations")
(count,) = cur.fetchone()
return count
def __contains__(self: SQLiteStore, key: str) -> bool:
"""Test whether the object contains the specified object or not."""
cur = self.con.cursor()
cur.execute("SELECT EXISTS(SELECT 1 FROM annotations WHERE [key] = ?)", (key,))
return cur.fetchone()[0] == 1
def __getitem__(self: SQLiteStore, key: str) -> Annotation:
"""Get an item from the store."""
cur = self.con.cursor()
cur.execute(
"""
SELECT geometry, properties, cx, cy
FROM annotations
WHERE [key] = :key
""",
{"key": key},
)
row = cur.fetchone()
if row is None:
raise KeyError(key)
serialised_geometry, serialised_properties, cx, cy = row
properties = json.loads(serialised_properties or "{}")
return Annotation(
properties=properties,
wkb=self._unpack_wkb(serialised_geometry, cx, cy),
)
[docs]
def keys(self: SQLiteStore) -> Iterable[int]:
"""Return an iterable (usually generator) of all keys in the store.
Returns:
Iterable[str]:
An iterable of keys.
"""
yield from self
def __iter__(self: SQLiteStore) -> Iterator[int]:
"""Return an iterator for the given object."""
cur = self.con.cursor()
cur.execute(
"""
SELECT [key]
FROM annotations
""",
)
while True:
row = cur.fetchone()
if row is None:
break
yield row[0] # The key
[docs]
def values(self: SQLiteStore) -> Iterable[tuple[int, Annotation]]:
"""Return an iterable of all annotation in the store.
Returns:
Iterable[Annotation]:
An iterable of annotations.
"""
for _, value in self.items(): # noqa: PERF102
yield value
[docs]
def items(self: SQLiteStore) -> Iterable[tuple[int, Annotation]]:
"""Return iterable (generator) over key and annotations."""
cur = self.con.cursor()
cur.execute(
"""
SELECT [key], cx, cy, geometry, properties
FROM annotations
""",
)
while True:
row = cur.fetchone()
if row is None:
break
key, cx, cy, serialised_geometry, serialised_properties = row
if serialised_geometry is not None:
geometry = self._unpack_geometry(serialised_geometry, cx, cy)
else:
geometry = Point(cx, cy)
properties = json.loads(serialised_properties)
yield key, Annotation(geometry, properties)
[docs]
def patch_many(
self: SQLiteStore,
keys: Iterable[int],
geometries: Iterable[Geometry] | None = None,
properties_iter: Iterable[Properties] | None = None,
) -> None:
"""Bulk patch of annotations.
This may be more efficient than calling `patch` repeatedly
in a loop.
Args:
geometries (iter(Geometry)):
An iterable of geometries to update.
properties_iter (iter(dict)):
An iterable of properties to update.
keys (iter(str)):
An iterable of keys for each annotation to be updated.
"""
# Validate inputs
if not any([geometries, properties_iter]):
msg = "At least one of geometries or properties_iter must be given"
raise ValueError(
msg,
)
keys = list(keys)
geometries = list(geometries) if geometries else None
properties_iter = list(properties_iter) if properties_iter else None
self._validate_equal_lengths(keys, geometries, properties_iter)
properties_iter = properties_iter or ({} for _ in keys) # pragma: no branch
geometries = geometries or (None for _ in keys) # pragma: no branch
# Update the database
cur = self.con.cursor()
# Begin a transaction
if self.auto_commit:
cur.execute("BEGIN")
for key, geometry, properties in zip(keys, geometries, properties_iter):
# Annotation is not in DB:
if key not in self:
self._append(key, Annotation(geometry, properties), cur)
continue
# Annotation is in DB:
if geometry:
self._patch_geometry(key, geometry, cur)
if properties:
cur.execute(
"""
UPDATE annotations
SET properties = json_patch(properties, :properties)
WHERE [key] = :key
""",
{
"key": key,
"properties": json.dumps(properties, separators=(",", ":")),
},
)
if self.auto_commit:
self.con.commit()
def _patch_geometry(
self: SQLiteStore,
key: str,
geometry: Geometry,
cur: sqlite3.Cursor,
) -> None:
"""Patch a geometry in the database.
Update the geometry of the annotation with the given key but
leave the properties untouched.
Args:
key (str): The key of the annotation to patch.
geometry (Geometry): The new geometry.
cur (sqlite3.Cursor): The cursor to use.
"""
bounds = dict(zip(("min_x", "min_y", "max_x", "max_y"), geometry.bounds))
xy = dict(zip("xy", np.array(geometry.centroid.coords[0])))
query_parameters = dict(
**bounds,
**xy,
key=key,
geometry=self.serialise_geometry(geometry),
)
cur.execute(
"""
UPDATE rtree
SET min_x = :min_x, min_y = :min_y,
max_x = :max_x, max_y = :max_y
WHERE EXISTS
(SELECT 1
FROM annotations
WHERE rtree.id = annotations.id
AND annotations.key == :key);
""",
query_parameters,
)
cur.execute(
"""
UPDATE annotations
SET cx = :x, cy = :y, geometry = :geometry
WHERE [key] = :key
""",
query_parameters,
)
[docs]
def remove_many(self: SQLiteStore, keys: Iterable[str]) -> None:
"""Bulk removal of annotations by keys.
Args:
keys (iter(str)):
An iterable of keys for the annotation to be removed.
"""
cur = self.con.cursor()
if self.auto_commit:
cur.execute("BEGIN")
for key in keys:
cur.execute(
"""
DELETE
FROM rtree
WHERE EXISTS
(SELECT 1
FROM annotations
WHERE rtree.id = annotations.id
AND annotations.key == ?);
""",
(key,),
)
cur.execute(
"DELETE FROM annotations WHERE [key] = ?",
(key,),
)
if self.auto_commit:
self.con.commit()
def __setitem__(self: SQLiteStore, key: str, annotation: Annotation) -> None:
"""Implements a method to assign a value to an item."""
if key in self:
self.patch(key, annotation.geometry, annotation.properties)
return
self.append(annotation, key)
def _get_table_columns(self: SQLiteStore) -> list[str]:
"""Get a list of columns in the annotations table."""
cur = self.con.execute("PRAGMA table_info(annotations)")
return [row[1] for row in cur.fetchall()]
[docs]
def add_area_column(self: SQLiteStore, *, mk_index: bool = True) -> None:
"""Add a column to store the area of the geometry."""
cur = self.con.cursor()
cur.execute(
"""
ALTER TABLE annotations
ADD COLUMN area INTEGER NOT NULL DEFAULT 0
""",
)
cur.execute(
"""
UPDATE annotations
SET area = get_area(geometry, cx, cy)
""",
)
if mk_index:
self.create_index("area", '"area"')
self.con.commit()
self.table_columns.append("area")
[docs]
def remove_area_column(self: SQLiteStore) -> None:
"""Remove the area column from the store."""
if "area" in self.indexes():
self.drop_index("area")
cur = self.con.cursor()
cur.execute(
"""
ALTER TABLE annotations
DROP COLUMN area
""",
)
self.con.commit()
self.table_columns.remove("area")
[docs]
def to_dataframe(self: SQLiteStore) -> pd.DataFrame:
"""Converts AnnotationStore to :class:`pandas.DataFrame`."""
store_to_df = pd.DataFrame()
df_rows = (
{
"key": key,
"geometry": annotation.geometry,
"properties": annotation.properties,
}
for key, annotation in self.items()
)
store_to_df = pd.concat([store_to_df, pd.json_normalize(df_rows)])
return store_to_df.set_index("key")
[docs]
def features(self: SQLiteStore) -> Generator[dict[str, object], None, None]:
"""Return annotations as a list of geoJSON features.
Returns:
list:
List of features as dictionaries.
"""
return (
{
"type": "Feature",
"geometry": geometry2feature(annotation.geometry),
"properties": annotation.properties,
}
for annotation in self.values()
)
[docs]
def commit(self: SQLiteStore) -> None:
"""Commit any in-memory changes to disk."""
self.con.commit()
[docs]
def dump(self: SQLiteStore, fp: Path | str | IO) -> None:
"""Serialise a copy of the whole store to a file-like object.
Args:
fp(Path or str or IO):
A file path or file handle object for output to disk.
"""
if hasattr(fp, "write"):
fp = fp.name
target = sqlite3.connect(fp)
self.con.backup(target)
[docs]
def dumps(self: SQLiteStore) -> str:
"""Serialise and return a copy of store as a string or bytes.
Returns:
str or bytes:
The serialised store.
"""
return "\n".join(self.con.iterdump())
[docs]
def clear(self: SQLiteStore) -> None:
"""Remove all annotations from the store."""
cur = self.con.cursor()
cur.execute("DELETE FROM rtree")
cur.execute("DELETE FROM annotations")
if self.auto_commit:
self.con.commit()
[docs]
def create_index(
self: SQLiteStore,
name: str,
where: str | bytes,
*,
analyze: bool = True,
) -> None:
"""Create an SQLite expression index based on the provided predicate.
Note that an expression index will only be used if the query expression
(in the WHERE clause) exactly matches the expression used when creating
the index (excluding minor inconsequential changes such as
whitespace).
An SQLite expression indexes require SQLite version 3.9.0 or higher.
Args:
name (str):
Name of the index to create.
where:
The predicate used to create the index.
analyze (bool):
Whether to run the "ANALYZE" command after creating the
index.
"""
_, minor, _ = sqlite3.sqlite_version_info
if minor < 9: # noqa: PLR2004
msg = "Requires sqlite version 3.9.0 or higher."
raise OSError(msg)
cur = self.con.cursor()
if not isinstance(where, str):
msg = f"Invalid type for `where` ({type(where)})."
raise TypeError(msg)
sql_predicate = eval( # skipcq: PYL-W0123, # noqa: PGH001, S307
where,
SQL_GLOBALS,
)
cur.execute(f"CREATE INDEX {name} ON annotations({sql_predicate})")
if analyze:
cur.execute(f"ANALYZE {name}")
[docs]
def indexes(self: SQLiteStore) -> list[str]:
"""Return a list of the names of all indexes in the store.
Returns:
List[str]:
The list of index names.
"""
cur = self.con.cursor()
cur.execute("SELECT name FROM sqlite_master WHERE TYPE = 'index'")
return [row[0] for row in cur.fetchall()]
[docs]
def drop_index(self: SQLiteStore, name: str) -> None:
"""Drop an index from the store.
Args:
name (str):
The name of the index to drop.
"""
cur = self.con.cursor()
cur.execute(f"DROP INDEX {name}")
[docs]
def optimize(self: SQLiteStore, limit: int = 1000, *, vacuum: bool = True) -> None:
"""Optimize the database with VACUUM and ANALYZE.
Args:
vacuum (bool):
Whether to run VACUUM.
limit (int):
The approximate maximum number of rows to examine when
running ANALYZE. If zero or negative, not limit will be
used. For more information see
https://www.sqlite.org/pragma.html#pragma_analysis_limit.
"""
if vacuum:
self.con.execute("VACUUM")
# Cannot use parameterized statements with PRAGMA!
self.con.execute(f"PRAGMA analysis_limit = {int(limit)}")
self.con.execute("PRAGMA optimize")
[docs]
class DictionaryStore(AnnotationStore):
"""Pure python dictionary backed annotation store."""
def __init__(
self: DictionaryStore,
connection: Path | str | IO = ":memory:",
) -> None:
"""Initialize :class:`DictionaryStore`."""
super().__init__()
self._rows = {}
self.connection = connection
self.path = self._connection_to_path(connection)
if self.connection not in [None, ":memory:"] and self.path.exists():
for line in self._load_cases(
fp=self.connection,
string_fn=lambda fp: fp.splitlines(),
file_fn=lambda fp: fp.readlines(),
):
dictionary = json.loads(line)
key = dictionary.get("key", uuid.uuid4().hex)
geometry = feature2geometry(dictionary["geometry"])
properties = dictionary["properties"]
self.append(Annotation(geometry, properties), key=key)
[docs]
def append(
self: DictionaryStore,
annotation: Annotation,
key: str | None = None,
) -> str:
"""Insert a new annotation, returning the key.
Args:
annotation (Annotation):
The shapely annotation to insert.
key (str):
Optional. The unique key used to identify the annotation in the
store. If not given a new UUID4 will be generated and returned
instead.
Returns:
str:
The unique key of the newly inserted annotation.
"""
if not isinstance(annotation.geometry, (Polygon, Point, LineString)):
msg = "Invalid geometry type. Must be one of Point, LineString, Polygon."
raise TypeError(msg)
key = key or str(uuid.uuid4())
self._rows[key] = {"annotation": annotation}
return key
[docs]
def patch(
self: DictionaryStore,
key: str,
geometry: Geometry | None = None,
properties: dict[str, object] | None = None,
) -> None:
"""Patch an annotation at given key.
Partial update of an annotation. Providing only a geometry will update
the geometry and leave properties unchanged. Providing a properties
dictionary applies a patch operation to the properties. Only updating
the properties which are given and leaving the rest unchanged. To
completely replace an annotation use `__setitem__`.
Args:
key(str):
The key of the annotation to update.
geometry(Geometry):
The new geometry. If None, the geometry is not updated.
properties(dict):
A dictionary of properties to patch and their new values.
If None, the existing properties are not altered.
"""
if key not in self:
self.append(Annotation(geometry, properties), key)
return
existing = self[key]
geometry = geometry or existing.geometry
properties = properties or {}
new_properties = copy.deepcopy(existing.properties)
new_properties.update(properties)
self[key] = Annotation(geometry, new_properties)
[docs]
def remove(self: DictionaryStore, key: str) -> None:
"""Remove annotation from the store with its unique key.
Args:
key (str):
The key of the annotation to be removed.
"""
del self._rows[key]
def __getitem__(self: DictionaryStore, key: str) -> Annotation:
"""Get an item from the store."""
return self._rows[key]["annotation"]
def __setitem__(self: DictionaryStore, key: str, annotation: Annotation) -> None:
"""Implements a method to assign a value to an item."""
if key in self._rows:
self._rows[key]["annotation"] = annotation
self._rows[key] = {"annotation": annotation}
def __contains__(self: DictionaryStore, key: str) -> bool:
"""Test whether the object contains the specified object or not."""
return key in self._rows
[docs]
def items(self: DictionaryStore) -> Generator[tuple[str, Annotation], None, None]:
"""Return iterable (generator) over key and annotations."""
for key, row in self._rows.items():
yield key, row["annotation"]
def __len__(self: DictionaryStore) -> int:
"""Return the length of the instance attributes."""
return len(self._rows)
# flake8: noqa: A003
[docs]
@classmethod
def open(cls: type[DictionaryStore], fp: Path | str | IO) -> AnnotationStore:
"""Opens :class:`DictionaryStore` from file pointer or path."""
return cls.from_ndjson(fp)
[docs]
def commit(self: DictionaryStore) -> None:
"""Commit any in-memory changes to disk."""
if str(self.connection) == ":memory:":
logger.warning("In-memory store. Nothing to commit.", stacklevel=2)
return
if not self.path.exists():
self.path.touch()
self.dump(self.connection)
[docs]
def dump(self: DictionaryStore, fp: Path | str | IO) -> str:
"""Serialise a copy of the whole store to a file-like object.
Args:
fp(Path or str or IO):
A file path or file handle object for output to disk.
"""
return self.to_ndjson(fp)
[docs]
def dumps(self: DictionaryStore) -> str:
"""Serialise and return a copy of store as a string or bytes.
Returns:
str or bytes:
The serialised store.
"""
return self.to_ndjson()
[docs]
def close(self: DictionaryStore) -> None:
"""Closes :class:`DictionaryStore` from file pointer or path."""
duplicate_filter = DuplicateFilter()
logger.addFilter(duplicate_filter)
# Try to commit any changes if the file is still open.
with contextlib.suppress(ValueError):
self.commit()
logger.removeFilter(duplicate_filter)