jsonpath-rfc9535 2.0.0__tar.gz → 2.0.2__tar.gz

{jsonpath_rfc9535-2.0.0 → jsonpath_rfc9535-2.0.2}/CHANGELOG.md

@@ -1,5 +1,25 @@
 # Python JSONPath RFC 9535 Change Log
 
+## Version 2.0.2
+
+**Fixes**
+
+- Restored support for traversing `dict` and `list` subclasses when resolving a JSONPath expression. See [#32](https://github.com/jg-rp/python-jsonpath-rfc9535/issues/32).
+
+## Version 2.0.1
+
+**Fixes**
+
+- Fixed trailing comma detection in bracketed segments. Previously we failed to consume whitespace after a comma and before checking for `]`. See [#26](https://github.com/jg-rp/python-jsonpath-rfc9535/issues/26).
+
+- Fixed comparison operators `<`, `>`, `<=` and `>=` when comparing `true` and/or `false` with `1` and `0`. Previously we were leaking Python behavior. See [#27](https://github.com/jg-rp/python-jsonpath-rfc9535/issues/27).
+
+- Fixed `\u` escaped sequence decoding. Previously we would not recognize lower case `e` and `f` hex digits due to a typo in a regular expression. See [#24](https://github.com/jg-rp/python-jsonpath-rfc9535/issues/24).
+
+- Fixed `\u` escape sequence rejection of code points less than or equal to 0x1F. The spec requires us to reject literals of 0x1F or lower, not escape sequences. See [#24](https://github.com/jg-rp/python-jsonpath-rfc9535/issues/24).
+
+- Fixed the shorthand name selector to accept the full range of allow Unicode code points and reject names containing `-`. See [#25](https://github.com/jg-rp/python-jsonpath-rfc9535/issues/25).
+
 ## Version 2.0.0
 
 This release includes performance improvements and some breaking API changes. JSONPath syntax and semantics are unchanged.

{jsonpath_rfc9535-2.0.0 → jsonpath_rfc9535-2.0.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: jsonpath-rfc9535
-Version: 2.0.0
+Version: 2.0.2
 Summary: RFC 9535 - JSONPath: Query Expressions for JSON in Python
 Project-URL: Documentation, https://jg-rp.github.io/python-jsonpath-rfc9535/
 Project-URL: Issues, https://github.com/jg-rp/python-jsonpath-rfc9535/issues

{jsonpath_rfc9535-2.0.0 → jsonpath_rfc9535-2.0.2}/pyproject.toml

@@ -49,6 +49,7 @@ dependencies = [
     "sphinx_copybutton",
     "types-regex>=2026.8.31.20260901",
     "ghp-import",
+    "twine",
 ]
 
 [tool.hatch.envs.benchmark]

jsonpath_rfc9535-2.0.2/src/jsonpath_rfc9535/__about__.py

@@ -0,0 +1 @@
+__version__ = "2.0.2"

{jsonpath_rfc9535-2.0.0 → jsonpath_rfc9535-2.0.2}/src/jsonpath_rfc9535/_lexer.py

@@ -6,16 +6,14 @@ from .exceptions import JSONPathSyntaxError
 RE_FLOAT = re.compile(r"(:?-?[0-9]+\.[0-9]+(?:[eE][+-]?[0-9]+)?)|(-?[0-9]+[eE]-[0-9]+)")
 RE_INT = re.compile(r"-?[0-9]+")
 RE_INT_EXP = re.compile(r"-?[0-9]+[eE]\+?[0-9]+")
-RE_NAME = re.compile(r"[\u0080-\uFFFFa-zA-Z_][\u0080-\uFFFFa-zA-Z0-9_-]*")
+RE_NAME = re.compile(
+    r"[\u0080-\uD7FF\uE000-\U0010FFFFa-zA-Z_][\u0080-\u10FFFF\uE000-\U0010FFFFa-zA-Z0-9_]*"
+)
 RE_TRIVIA = re.compile(r"[ \n\r\t]+")
 
 
 def tokenize(source: str) -> list[Token]:
-    """Transform a JSONPath query source string into a list of tokens.
-
-    This tokenizer emits some non-standard tokens. The parser is responsible
-    for reject non-standard tokens with a syntax error if desired.
-    """
+    """Transform a JSONPath query source string into a list of tokens."""
     tokens: list[Token] = []
     length = len(source)
     pos = 0
@@ -238,7 +236,8 @@ def _is_name_first_ch(ch: int) -> bool:
         (ch >= 65 and ch <= 90)
         or (ch >= 97 and ch <= 122)
         or ch == 95
-        or (ch >= 0x80 and ch <= 0xFFFF)
+        or (ch >= 0x80 and ch <= 0xD7FF)
+        or (ch >= 0xE000 and ch <= 0x10FFFF)
     )
 
 

{jsonpath_rfc9535-2.0.0 → jsonpath_rfc9535-2.0.2}/src/jsonpath_rfc9535/_parser.py

@@ -20,7 +20,7 @@ from .query import JSONPathQuery
 if TYPE_CHECKING:
     from ._environment import JSONPathEnvironment
 
-_RE_ESCAPE_U = re.compile(r"(\\u[0-9a-dA-F]{4}|\\.)")
+_RE_ESCAPE_U = re.compile(r"(\\u[0-9a-fA-F]{4}|\\.)")
 
 
 class Parser(Protocol):
@@ -344,13 +344,6 @@ class Parser_(ABC):
                         ((code_point & 0x03FF) << 10) | (low_surrogate & 0x03FF)
                     )
 
-                if code_point <= 0x1F:
-                    raise JSONPathSyntaxError(
-                        "invalid escape sequence",
-                        token,
-                        self.source,
-                    )
-
                 unescaped.append(chr(code_point))
             else:
                 raise JSONPathSyntaxError(

{jsonpath_rfc9535-2.0.0 → jsonpath_rfc9535-2.0.2}/src/jsonpath_rfc9535/_parser_standard.py

@@ -195,6 +195,7 @@ class StandardParser(Parser_):
                 )
 
             self.eat(TOKEN_COMMA)
+            self.skip(TOKEN_TRIVIA)
             if self.kind() == TOKEN_RIGHT_BRACKET:
                 raise JSONPathSyntaxError(
                     "unexpected trailing comma",

{jsonpath_rfc9535-2.0.0 → jsonpath_rfc9535-2.0.2}/src/jsonpath_rfc9535/_resolve_standard.py

@@ -66,37 +66,37 @@ class StandardResolver(Resolver_):
 
         match selector:
             case (_ast.NAME_SELECTOR, _, name):
-                if type(obj) is dict and name in obj:
+                if isinstance(obj, dict) and name in obj:
                     yield self.new_child(node, name, obj[name])
 
             case (_ast.INDEX_SELECTOR, _, index):
-                if type(obj) is list and len(obj) >= abs(index) + int(index >= 0):
+                if isinstance(obj, list) and len(obj) >= abs(index) + int(index >= 0):
                     if index < 0 and len(obj) >= abs(index):
                         index = len(obj) + index
 
                     yield self.new_child(node, index, obj[index])
 
             case (_ast.SLICE_SELECTOR, _, slice_):
-                if type(obj) is list and slice_.step != 0:
+                if isinstance(obj, list) and slice_.step != 0:
                     for i, elem in zip(range(*slice_.indices(len(obj))), obj[slice_]):
                         yield self.new_child(node, i, elem)
 
             case (_ast.WILDCARD_SELECTOR, _):
-                if type(obj) is dict:
+                if isinstance(obj, dict):
                     for k, v in obj.items():
                         yield self.new_child(node, k, v)
 
-                elif type(obj) is list:
+                elif isinstance(obj, list):
                     for i, elem in enumerate(obj):
                         yield self.new_child(node, i, elem)
 
             case (_ast.FILTER_SELECTOR, _, expr):
-                if type(obj) is dict:
+                if isinstance(obj, dict):
                     for k, v in obj.items():
                         if self.truthy(self._evaluate_expression(expr, k, v)):
                             yield self.new_child(node, k, v)
 
-                elif type(obj) is list:
+                elif isinstance(obj, list):
                     for i, elem in enumerate(obj):
                         if self.truthy(self._evaluate_expression(expr, i, elem)):
                             yield self.new_child(node, i, elem)
@@ -210,16 +210,14 @@ class StandardResolver(Resolver_):
             yield node
             obj = node[0]
 
-            if type(obj) is dict:
+            if isinstance(obj, dict):
                 for k, v in obj.items():
-                    vt = type(v)
-                    if isinstance(k, str) and (vt is list or vt is dict):
+                    if isinstance(k, str) and isinstance(v, (dict, list)):
                         yield from visit_(self.new_child(node, k, v), depth + 1)
 
-            elif type(obj) is list:
+            elif isinstance(obj, list):
                 for i, elem in enumerate(obj):
-                    et = type(elem)
-                    if et is list or et is dict:
+                    if isinstance(elem, (dict, list)):
                         yield from visit_(self.new_child(node, i, elem), depth + 1)
 
         return visit_(node, 1)

{jsonpath_rfc9535-2.0.0 → jsonpath_rfc9535-2.0.2}/src/jsonpath_rfc9535/_resolver.py

@@ -105,6 +105,9 @@ class Resolver_:
             else:
                 return False
 
+        if isinstance(left, bool) or isinstance(right, bool):
+            return False
+
         if isinstance(left, str) and isinstance(right, str):
             return left < right
 

{jsonpath_rfc9535-2.0.0 → jsonpath_rfc9535-2.0.2}/src/jsonpath_rfc9535/_resolver_basic.py

@@ -64,37 +64,37 @@ class BasicResolver(Resolver_):
     def _resolve_selector(self, selector: Selector, obj: object) -> Iterable[object]:
         match selector:
             case (_ast.NAME_SELECTOR, _, name):
-                if type(obj) is dict and name in obj:
+                if isinstance(obj, dict) and name in obj:
                     yield obj[name]
 
             case (_ast.INDEX_SELECTOR, _, index):
-                if type(obj) is list and len(obj) >= abs(index) + int(index >= 0):
+                if isinstance(obj, list) and len(obj) >= abs(index) + int(index >= 0):
                     if index < 0 and len(obj) >= abs(index):
                         index = len(obj) + index
 
                     yield obj[index]
 
             case (_ast.SLICE_SELECTOR, _, slice_):
-                if type(obj) is list and slice_.step != 0:
+                if isinstance(obj, list) and slice_.step != 0:
                     for i, elem in zip(range(*slice_.indices(len(obj))), obj[slice_]):
                         yield elem
 
             case (_ast.WILDCARD_SELECTOR, _):
-                if type(obj) is dict:
+                if isinstance(obj, dict):
                     for k, v in obj.items():
                         yield v
 
-                elif type(obj) is list:
+                elif isinstance(obj, list):
                     for i, elem in enumerate(obj):
                         yield elem
 
             case (_ast.FILTER_SELECTOR, _, expr):
-                if type(obj) is dict:
+                if isinstance(obj, dict):
                     for k, v in obj.items():
                         if self.truthy(self._evaluate_expression(expr, k, v)):
                             yield v
 
-                elif type(obj) is list:
+                elif isinstance(obj, list):
                     for i, elem in enumerate(obj):
                         if self.truthy(self._evaluate_expression(expr, i, elem)):
                             yield elem
@@ -210,16 +210,14 @@ class BasicResolver(Resolver_):
 
             yield obj
 
-            if type(obj) is dict:
+            if isinstance(obj, dict):
                 for k, v in obj.items():
-                    vt = type(v)
-                    if isinstance(k, str) and (vt is list or vt is dict):
+                    if isinstance(k, str) and isinstance(v, (list, dict)):
                         yield from visit_(v, depth + 1)
 
-            elif type(obj) is list:
+            elif isinstance(obj, list):
                 for elem in obj:
-                    et = type(elem)
-                    if et is list or et is dict:
+                    if isinstance(elem, (dict, list)):
                         yield from visit_(elem, depth + 1)
 
         return visit_(node, 1)

jsonpath_rfc9535-2.0.2/tests/test_issues.py

@@ -0,0 +1,86 @@
+import pytest
+
+import jsonpath_rfc9535 as jsonpath
+
+
+def test_issue_13() -> None:
+    # This was failing with "unbalanced parentheses".
+    _q = jsonpath.compile("$[? count(@.likes[? @.location]) > 3]")
+
+
+def test_issue_21() -> None:
+    data = {"foo": {"bar": {"baz": 42}}}
+    node = jsonpath.find_one("$.foo.bar.baz", data)
+
+    expected = 42
+    assert node is not None
+    assert node.value == expected
+    assert data["foo"]["bar"]["baz"] == expected
+
+    new_value = 99
+    node.value = new_value
+    assert node.value == new_value
+    assert data["foo"]["bar"]["baz"] == new_value
+
+    parent = node.parent()
+    assert parent is not None
+    assert parent.value == {"baz": new_value}
+    assert parent.value["baz"] == new_value  # type: ignore
+
+
+def test_issue_24() -> None:
+    data = {"\u001f": 1, "é": 2}
+
+    assert jsonpath.findall("$['\\u001f']", data) == [1]
+    assert jsonpath.findall("$['\\u00e9']", data) == [2]
+    assert jsonpath.findall("$['\\u00E9']", data) == [2]
+
+
+def test_issue_25() -> None:
+    with pytest.raises(jsonpath.JSONPathSyntaxError):
+        assert jsonpath.compile("$.a-b")
+
+    assert jsonpath.findall("$.😀", {"😀": 1}) == [1]
+    assert jsonpath.findall("$['😀']", {"😀": 1}) == [1]
+
+
+def test_issue_26() -> None:
+    with pytest.raises(jsonpath.JSONPathSyntaxError):
+        jsonpath.compile("$[1,]")
+
+    with pytest.raises(jsonpath.JSONPathSyntaxError):
+        jsonpath.compile("$[1, ]")
+
+    with pytest.raises(jsonpath.JSONPathSyntaxError):
+        jsonpath.compile("$[ 1, ]")
+
+
+def test_issue_27() -> None:
+    assert jsonpath.findall("$[?@ > false]", [0, 1, True, False]) == []
+    assert jsonpath.findall("$[?@ < true]", [0, 1, True]) == []
+    assert jsonpath.findall("$[?@ >= 0]", [True, False, 0]) == [0]
+
+
+def test_issue_32() -> None:
+    class MyDict(dict[str, object]):
+        pass
+
+    class MyList(list[object]):
+        pass
+
+    data = MyDict(
+        {
+            "users": MyList(
+                [
+                    {"id": "usr_101", "name": "Alice", "status": "pending"},
+                    {"id": "usr_102", "name": "Bob", "status": "active"},
+                    {"id": "usr_103", "name": "Charlie", "status": "pending"},
+                ]
+            ),
+        }
+    )
+
+    expr = "$.users[?@.status == 'pending'].name"
+
+    assert jsonpath.findall(expr, data) == ["Alice", "Charlie"]
+    assert jsonpath.find(expr, data).values() == ["Alice", "Charlie"]

jsonpath_rfc9535-2.0.0/src/jsonpath_rfc9535/__about__.py

@@ -1 +0,0 @@
-__version__ = "2.0.0"

jsonpath_rfc9535-2.0.0/tests/test_issues.py

@@ -1,26 +0,0 @@
-import jsonpath_rfc9535 as jsonpath
-
-
-def test_issue_13() -> None:
-    # This was failing with "unbalanced parentheses".
-    _q = jsonpath.compile("$[? count(@.likes[? @.location]) > 3]")
-
-
-def test_issue_21() -> None:
-    data = {"foo": {"bar": {"baz": 42}}}
-    node = jsonpath.find_one("$.foo.bar.baz", data)
-
-    expected = 42
-    assert node is not None
-    assert node.value == expected
-    assert data["foo"]["bar"]["baz"] == expected
-
-    new_value = 99
-    node.value = new_value
-    assert node.value == new_value
-    assert data["foo"]["bar"]["baz"] == new_value
-
-    parent = node.parent()
-    assert parent is not None
-    assert parent.value == {"baz": new_value}
-    assert parent.value["baz"] == new_value  # type: ignore