JSONPath Sleuth

Fast Python bindings (via Rust + PyO3) for:

• Resolving JSONPath expressions against Python dict/list JSON
• Finding JSONPath-like paths for all occurrences of a target value
• Extracting all JSONPath-like paths paired with their leaf values

Install / Build

• Option A: editable dev install
  • pip install maturin
  • maturin develop -m pyproject.toml
• Option B: build wheel
  • maturin build -m pyproject.toml
  • pip install dist/*.whl

Requires Python 3.8+.

Run Tests

• Rust unit tests (no Python needed)

  • cargo test --no-default-features
  • Also compile PyO3 bindings: PYO3_PYTHON=$(which python3) cargo test --features python

• Python tests (pytest)

  • Option A (quick):
    • python3 -m venv .venv && source .venv/bin/activate
    • pip install -U pip pytest maturin
    • maturin develop -m Cargo.toml --features python
    • pytest -q
  • Option B (dev extra):
    • python3 -m venv .venv && source .venv/bin/activate
    • pip install -U pip
    • pip install -e .[dev]
    • maturin develop -m Cargo.toml --features python
    • pytest -q

Notes

• Re-run maturin develop after Rust changes to refresh the extension in your venv.
• If pytest cannot import jsonpath_sleuth, ensure you activated the same venv used for maturin develop.

Publish

• Build wheels + sdist

  • maturin build -m Cargo.toml --features python --release --sdist

• TestPyPI (requires separate TestPyPI account and token)

  • Publish: maturin publish -m Cargo.toml --features python --repository-url https://test.pypi.org/legacy/ -u __token__ -p <pypi-TEST_TOKEN>
  • Install to verify: pip install -i https://test.pypi.org/simple jsonpath-sleuth

• PyPI

  • Publish: maturin publish -m Cargo.toml --features python -u __token__ -p <pypi-PROD_TOKEN>
  • Install to verify: pip install jsonpath-sleuth

Tips

• Bump version in both Cargo.toml and pyproject.toml before publishing a new release.
• Tokens begin with pypi-. Avoid committing tokens; pass on the command line or configure ~/.pypirc.

Python API

Module: jsonpath_sleuth

• resolve_jsonpath(data: dict | list, path: str) -> list[Any]
  • Returns a list of matched values for the given JSONPath. The path may omit the leading $ (it is added automatically).
• find_jsonpaths_by_value(data: dict | list, target: Any) -> list[str]
  • Returns string paths like foo.bar[0].baz where value equals target.
• extract_jsonpaths_and_values(data: dict | list) -> list[tuple[str, Any]]
  • Returns all JSONPath-like paths paired with their leaf values. Paths use . for object keys and [idx] for arrays.

Examples

from jsonpath_sleuth import resolve_jsonpath, find_jsonpaths_by_value, extract_jsonpaths_and_values

obj = {
    "store": {
        "book": [
            {"category": "fiction", "title": "Sword"},
            {"category": "fiction", "title": "Shield"},
        ],
        "bicycle": {"color": "red", "price": 19.95},
    }
}

# 1) Resolve JSONPath (prefix not required)
print(resolve_jsonpath(obj, "store.book[*].title"))
# -> ["Sword", "Shield"]
# Also works with explicit JSONPath:
print(resolve_jsonpath(obj, "$.store.book[*].title"))
# -> ["Sword", "Shield"]

# 2) Find paths by target value
print(find_jsonpaths_by_value(obj, "fiction"))
# -> ["store.book[0].category", "store.book[1].category"]

# 3) Extract all leaf paths and values
print(extract_jsonpaths_and_values(obj))
# -> [
#     ("store.book[0].category", "fiction"),
#     ("store.book[0].title", "Sword"),
#     ("store.book[1].category", "fiction"),
#     ("store.book[1].title", "Shield"),
#     ("store.bicycle.color", "red"),
#     ("store.bicycle.price", 19.95),
# ]

Advanced: Nested Wildcard Filters

JSONPath Sleuth supports nested wildcard filters - a powerful feature that most JSONPath libraries don't handle well:

from jsonpath_sleuth import resolve_jsonpath

data = {
    "parties": [
        {
            "name": "Alice",
            "results": [
                {"item": "A"},
                {"item": "B"}
            ]
        },
        {
            "name": "Bob",
            "results": []
        },
        {
            "name": "Charlie",
            "results": [
                {"item": "A"},
                {"item": "C"}
            ]
        }
    ]
}

# Find all parties that have ANY result with item='A'
print(resolve_jsonpath(data, "parties[?(@.results[*].item=='A')].name"))
# -> ["Alice", "Charlie"]

# This pattern works: <base>[?(@.<nested_array>[*].<field>=='<value>')].<result_field>

How it works:

• Checks if ANY item in the nested array matches the condition
• Returns the specified field from matching parent objects
• Custom implementation handles what standard JSONPath libraries can't

Notes

JSONPath Support

• Standard JSONPath is powered by jsonpath-rust crate
• Nested wildcard filters use custom implementation for enhanced functionality
• JSONPath keys with spaces or special characters must be quoted using bracket notation
  • Example: use a['some key'].next instead of a.some key.next
  • You may omit the leading $; the resolver adds it automatically

Path Format

• Paths produced by value search use . between object keys and [idx] for arrays
• If the entire input equals the target, no paths are returned (empty list)

Supported Patterns

• ✅ Standard wildcards: store.book[*].title
• ✅ Filters: store.book[?(@.price < 10)].title
• ✅ Nested wildcard filters: parties[?(@.results[*].item=='A')].name
• ✅ Recursive descent: $..price
• ✅ Array slices: store.book[0:2].title