"""A fluent API for working with `JSONPathMatch` iterators.""" from __future__ import annotations import collections import itertools from enum import Enum from enum import auto from typing import TYPE_CHECKING from typing import Any from typing import Dict from typing import Iterable from typing import Iterator from typing import List from typing import Mapping from typing import Optional from typing import Sequence from typing import Tuple from typing import Union if TYPE_CHECKING: from jsonpath import CompoundJSONPath from jsonpath import JSONPath from jsonpath import JSONPathEnvironment from jsonpath import JSONPathMatch from jsonpath import JSONPointer class Projection(Enum): """Projection style used by `Query.select()`.""" RELATIVE = auto() """The default projection. Selections include parent arrays and objects relative to the JSONPathMatch.""" ROOT = auto() """Selections include parent arrays and objects relative to the root JSON value.""" FLAT = auto() """All selections are appended to a new array/list, without arrays and objects on the path to the selected value.""" class Query: """A fluent API for managing `JSONPathMatch` iterators. Usually you'll want to use `jsonpath.query()` or `JSONPathEnvironment.query()` to create instances of `Query` rather than instantiating `Query` directly. Arguments: it: A `JSONPathMatch` iterable, as you'd get from `jsonpath.finditer()` or `JSONPathEnvironment.finditer()`. **New in version 1.1.0** """ def __init__(self, it: Iterable[JSONPathMatch], env: JSONPathEnvironment) -> None: self._it = iter(it) self._env = env def __iter__(self) -> Iterator[JSONPathMatch]: return self._it def limit(self, n: int) -> Query: """Limit the query iterator to at most _n_ matches. Raises: ValueError: If _n_ < 0. """ if n < 0: raise ValueError("can't limit by a negative number of matches") self._it = itertools.islice(self._it, n) return self def head(self, n: int) -> Query: """Limit the query iterator to at most the first _n_ matches. `head()` is an alias for `limit()`. Raises: ValueError: If _n_ < 0. """ return self.limit(n) def first(self, n: int) -> Query: """Limit the query iterator to at most the first _n_ matches. `first()` is an alias for `limit()`. Raises: ValueError: If _n_ < 0. """ return self.limit(n) def drop(self, n: int) -> Query: """Skip up to _n_ matches from the query iterator. Raises: ValueError: If _n_ < 0. """ if n < 0: raise ValueError("can't drop a negative number of matches") if n > 0: next(itertools.islice(self._it, n, n), None) return self def skip(self, n: int) -> Query: """Skip up to _n_ matches from the query iterator. Raises: ValueError: If _n_ < 0. """ return self.drop(n) def tail(self, n: int) -> Query: """Drop matches up to the last _n_ matches from the iterator. Raises: ValueError: If _n_ < 0. """ if n < 0: raise ValueError("can't select a negative number of matches") self._it = iter(collections.deque(self._it, maxlen=n)) return self def last(self, n: int) -> Query: """Drop up to the last _n_ matches from the iterator. `last()` is an alias for `tail()`. Raises: ValueError: If _n_ < 0. """ return self.tail(n) def values(self) -> Iterable[object]: """Return an iterable of objects associated with each match.""" return (m.obj for m in self._it) def locations(self) -> Iterable[str]: """Return an iterable of normalized paths, one for each match.""" return (m.path for m in self._it) def items(self) -> Iterable[Tuple[str, object]]: """Return an iterable of (path, object) tuples, one for each match.""" return ((m.path, m.obj) for m in self._it) def pointers(self) -> Iterable[JSONPointer]: """Return an iterable of JSONPointers, one for each match.""" return (m.pointer() for m in self._it) def first_one(self) -> Optional[JSONPathMatch]: """Return the first `JSONPathMatch` or `None` if there were no matches.""" try: return next(self._it) except StopIteration: return None def one(self) -> Optional[JSONPathMatch]: """Return the first `JSONPathMatch` or `None` if there were no matches. `one()` is an alias for `first_one()`. """ return self.first_one() def last_one(self) -> Optional[JSONPathMatch]: """Return the last `JSONPathMatch` or `None` if there were no matches.""" try: return next(iter(self.tail(1))) except StopIteration: return None def tee(self, n: int = 2) -> Tuple[Query, ...]: """Return _n_ independent queries by teeing this query's iterator. It is not safe to use a `Query` instance after calling `tee()`. """ return tuple(Query(it, self._env) for it in itertools.tee(self._it, n)) def take(self, n: int) -> Query: """Return a new query iterating over the next _n_ matches. It is safe to continue using this query after calling take. """ return Query(list(itertools.islice(self._it, n)), self._env) def select( self, *expressions: Union[str, JSONPath, CompoundJSONPath], projection: Projection = Projection.RELATIVE, ) -> Iterable[object]: """Query projection using relative JSONPaths. Arguments: expressions: One or more JSONPath query expressions to select relative to each match in this query iterator. projection: The style of projection used when selecting values. Can be one of `Projection.RELATIVE`, `Projection.ROOT` or `Projection.FLAT`. Defaults to `Projection.RELATIVE`. Returns: An iterable of objects built from selecting _expressions_ relative to each match from the current query. **New in version 1.2.0** """ return filter( bool, (self._select(m, expressions, projection) for m in self._it), ) def _select( self, match: JSONPathMatch, expressions: Tuple[Union[str, JSONPath, CompoundJSONPath], ...], projection: Projection, ) -> object: if not isinstance(match.obj, (Mapping, Sequence)) or isinstance(match.obj, str): return None if projection == Projection.RELATIVE: obj: Dict[Union[int, str], Any] = {} for expr in expressions: path = self._env.compile(expr) if isinstance(expr, str) else expr for rel_match in path.finditer(match.obj): # type: ignore _patch_obj(rel_match.parts, obj, rel_match.obj) return _fix_sparse_arrays(obj) if projection == Projection.FLAT: arr: List[object] = [] for expr in expressions: path = self._env.compile(expr) if isinstance(expr, str) else expr for rel_match in path.finditer(match.obj): # type: ignore arr.append(rel_match.obj) return arr # Project from the root document obj = {} for expr in expressions: path = self._env.compile(expr) if isinstance(expr, str) else expr for rel_match in path.finditer(match.obj): # type: ignore _patch_obj(match.parts + rel_match.parts, obj, rel_match.obj) return _fix_sparse_arrays(obj) def _patch_obj( parts: Tuple[Union[int, str], ...], obj: Mapping[Union[str, int], Any], value: object, ) -> None: _obj = obj # For lack of a better idea, we're patching arrays to dictionaries with # integer keys. This is to handle sparse array selections without having # to keep track of indexes and how they map from the root JSON value to # the selected JSON value. # # We'll fix these "sparse arrays" after the patch has been applied. for part in parts[:-1]: if part not in _obj: _obj[part] = {} # type: ignore _obj = _obj[part] _obj[parts[-1]] = value # type: ignore def _fix_sparse_arrays(obj: Any) -> object: """Fix sparse arrays (dictionaries with integer keys).""" if isinstance(obj, str) or not obj: return obj if isinstance(obj, Sequence): return [_fix_sparse_arrays(e) for e in obj] if isinstance(obj, Mapping): if isinstance(next(iter(obj)), int): return [_fix_sparse_arrays(v) for v in obj.values()] return {k: _fix_sparse_arrays(v) for k, v in obj.items()} return obj