A final implementation of JSONPath for Python that aims to be standard compliant, including arithmetic and binary comparison operators, as defined in the original JSONPath proposal.
This packages merges both jsonpath-rw and jsonpath-rw-ext and provides several AST API enhancements, such as the ability to update or remove nodes in the tree.
This library provides a robust and significantly extended implementation of JSONPath for Python. It is tested with CPython 3.8 and higher.
This library differs from other JSONPath implementations in that it is a full language implementation, meaning the JSONPath expressions are first class objects, easy to analyze, transform, parse, print, and extend.
To install, use pip:
$ pip install --upgrade jsonpath-ngBasic examples:
$ python
>>> from jsonpath_ng import jsonpath, parse
# A robust parser, not just a regex. (Makes powerful extensions possible; see below)
>>> jsonpath_expr = parse('foo[*].baz')
# Extracting values is easy
>>> [match.value for match in jsonpath_expr.find({'foo': [{'baz': 1}, {'baz': 2}]})]
[1, 2]
# Matches remember where they came from
>>> [str(match.full_path) for match in jsonpath_expr.find({'foo': [{'baz': 1}, {'baz': 2}]})]
['foo.[0].baz', 'foo.[1].baz']
# Modifying values matching the path
>>> jsonpath_expr.update( {'foo': [{'baz': 1}, {'baz': 2}]}, 3)
{'foo': [{'baz': 3}, {'baz': 3}]}
# Modifying one of the values matching the path
>>> matches = jsonpath_expr.find({'foo': [{'baz': 1}, {'baz': 2}]})
>>> matches[0].full_path.update( {'foo': [{'baz': 1}, {'baz': 2}]}, 3)
{'foo': [{'baz': 3}, {'baz': 2}]}
# Removing all values matching a path
>>> jsonpath_expr.filter(lambda d: True, {'foo': [{'baz': 1}, {'baz': 2}]})
{'foo': [{}, {}]}
# Removing values containing particular data matching path
>>> jsonpath_expr.filter(lambda d: d == 2, {'foo': [{'baz': 1}, {'baz': 2}]})
{'foo': [{'baz': 1}, {}]}
# And this can be useful for automatically providing ids for bits of data that do not have them (currently a global switch)
>>> jsonpath.auto_id_field = 'id'
>>> [match.value for match in parse('foo[*].id').find({'foo': [{'id': 'bizzle'}, {'baz': 3}]})]
['foo.bizzle', 'foo.[1]']
# A handy extension: named operators like `parent`
>>> [match.value for match in parse('a.*.b.`parent`.c').find({'a': {'x': {'b': 1, 'c': 'number one'}, 'y': {'b': 2, 'c': 'number two'}}})]
['number two', 'number one']
# You can also build expressions directly quite easily
>>> from jsonpath_ng.jsonpath import Fields
>>> from jsonpath_ng.jsonpath import Slice
>>> jsonpath_expr_direct = Fields('foo').child(Slice('*')).child(Fields('baz')) # This is equivalentUsing the extended parser:
$ python
>>> from jsonpath_ng.ext import parse
# A robust parser, not just a regex. (Makes powerful extensions possible; see below)
>>> jsonpath_expr = parse('foo[*].baz')The JSONPath syntax supported by this library includes some additional
features and omits some problematic features (those that make it
unportable). In particular, some new operators such as | and
where are available, and parentheses are used for grouping not for
callbacks into Python, since with these changes the language is not
trivially associative. Also, fields may be quoted whether or not they
are contained in brackets.
Atomic expressions:
| Syntax | Meaning |
|---|---|
$ |
The root object |
`this` |
The "current" object. |
`foo` |
More generally, this syntax allows "named operators" to extend JSONPath is arbitrary ways |
| field | Specified field(s), described below |
[ field ] |
Same as field |
[ idx ] |
Array access, described below (this is always unambiguous with field access) |
Jsonpath operators:
| Syntax | Meaning |
|---|---|
jsonpath1 . jsonpath2 |
All nodes matched by jsonpath2 starting at any node matching jsonpath1 |
jsonpath [ whatever ] |
Same as jsonpath.whatever |
jsonpath1 .. jsonpath2 |
All nodes matched by jsonpath2 that descend from any node matching jsonpath1 |
jsonpath1 where jsonpath2 |
Any nodes matching jsonpath1 with a child matching jsonpath2 |
jsonpath1 wherenot jsonpath2 |
Any nodes matching jsonpath1 with a child not matching jsonpath2 |
jsonpath1 | jsonpath2 |
Any nodes matching the union of jsonpath1 and jsonpath2 |
Field specifiers ( field ):
| Syntax | Meaning |
|---|---|
fieldname |
the field fieldname (from the "current" object) |
"fieldname" |
same as above, for allowing special characters in the fieldname |
'fieldname' |
ditto |
* |
any field |
field , field |
either of the named fields (you can always build equivalent jsonpath using |) |
Array specifiers ( idx ):
| Syntax | Meaning |
|---|---|
[n] |
array index (may be comma-separated list) |
[start?:end?] |
array slicing (note that step is unimplemented only due to lack of need thus far) |
[*] |
any array index |
If you are programming in Python and would like a more robust way to create JSONPath expressions that does not depend on a parser, it is very easy to do so directly, and here are some examples:
Root()Slice(start=0, end=None, step=None)Fields('foo', 'bar')Index(42)Child(Fields('foo'), Index(42))Where(Slice(), Fields('subfield'))Descendants(jsonpath, jsonpath)
- Path data: The result of
JsonPath.findprovide detailed context and path data so it is easy to traverse to parent objects, print full paths to pieces of data, and generate automatic ids. - Automatic Ids: If you set
jsonpath_ng.auto_id_fieldto a value other than None, then for any piece of data missing that field, it will be replaced by the JSONPath to it, giving automatic unique ids to any piece of data. These ids will take into account any ids already present as well. - Named operators: Instead of using
@to reference the current object, this library uses`this`. In general, any string contained in backquotes can be made to be a new operator, currently by extending the library.
To use the extensions below you must import from jsonpath_ng.ext.
| name | Example |
|---|---|
| len |
|
| sub |
|
| split |
|
| sorted |
|
| filter |
Supported operators: - Equality: ==, =, != - Comparison: >, >=, <, <= - Regex match: =~ Combine multiple criteria with '&'. Properties can only be compared to static values. |
| arithmetic (-+*/) |
|
Operations are done with python operators and allows types that python allows, and return [] if the operation can be done due to incompatible types.
When operators are used, a jsonpath must be be fully defined otherwise jsonpath-rw-ext can't known if the expression is a string or a jsonpath field, in this case it will choice string as type.
Example with data:
{
'cow': 'foo',
'fish': 'bar'
}
cow + fish returns cowfish$.cow + $.fish returns foobar$.cow + "_" + $.fish returns foo_bar$.cow + "_" + fish returns foo_fishArithmetic can be used against two lists if they have the same size.
Example with data:
{'objects': [
{'cow': 2, 'cat': 3},
{'cow': 4, 'cat': 6}
]}
$.objects[\*].cow + $.objects[\*].cat returns [6, 9]There are way too many JSONPath implementations out there to discuss. Some are robust, some are toy projects that still work fine, some are exercises. There will undoubtedly be many more. This one is made for use in released, maintained code, and in particular for programmatic access to the abstract syntax and extension. But JSONPath at its simplest just isn't that complicated, so you can probably use any of them successfully. Why not this one?
The original proposal, as far as I know:
- JSONPath - XPath for JSON by Stefan Goessner.
Loading json data from file
import json
d = json.loads('{"foo": [{"baz": 1}, {"baz": 2}]}')
# or
with open('myfile.json') as f:
d = json.load(f)The main parsing toolkit underlying this library,
PLY, does not work with docstrings
removed. For example, PYTHONOPTIMIZE=2 and python -OO will both
cause a failure.
This package is authored and maintained by:
- Kenn Knowles (@kennknowles)
- Tomas Aparicio <https://github.com/h2non>
with the help of patches submitted by these contributors.
Copyright 2013 - Kenneth Knowles
Copyright 2017 - Tomas Aparicio
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.