hat.json

18def equals(a: Data,
19           b: Data
20           ) -> bool:
21    """Equality comparison of json serializable data.
22
23    Tests for equality of data according to JSON format. Notably, ``bool``
24    values are not considered equal to numeric values in any case. This is
25    different from default equality comparison, which considers `False`
26    equal to `0` and `0.0`; and `True` equal to `1` and `1.0`.
27
28    Example::
29
30        assert equals(0, 0.0) is True
31        assert equals({'a': 1, 'b': 2}, {'b': 2, 'a': 1}) is True
32        assert equals(1, True) is False
33
34    """
35    if a is None:
36        return b is None
37
38    if isinstance(a, bool):
39        return isinstance(b, bool) and a == b
40
41    if isinstance(a, (int, float)):
42        return (isinstance(b, (int, float)) and
43                not isinstance(b, bool) and
44                a == b)
45
46    if isinstance(a, str):
47        return isinstance(b, str) and a == b
48
49    if isinstance(a, list):
50        return (isinstance(b, list) and
51                len(a) == len(b) and
52                all(equals(i, j) for i, j in zip(a, b)))
53
54    if isinstance(a, dict):
55        return (isinstance(b, dict) and
56                len(a) == len(b) and
57                all(equals(a[key], b[key]) for key in a.keys()))
58
59    raise TypeError('invalid json type')

62def clone(data: Data) -> Data:
63    """Deep clone data
64
65    This function recursively creates new instances of array and object data
66    based on input data. Resulting json data is equal to provided data.
67
68    Example::
69
70        x = {'a': [1, 2, 3]}
71        y = clone(x)
72        assert x is not y
73        assert x['a'] is not y['a']
74        assert equals(x, y)
75
76    """
77    if isinstance(data, list):
78        return [clone(i) for i in data]
79
80    if isinstance(data, dict):
81        return {k: clone(v) for k, v in data.items()}
82
83    return data

 86def flatten(data: Data) -> Iterable[Data]:
 87    """Flatten JSON data
 88
 89    If `data` is array, this generator recursively yields result of `flatten`
 90    call with each element of input list. For other `Data` types, input data is
 91    yielded.
 92
 93    Example::
 94
 95        data = [1, [], [2], {'a': [3]}]
 96        result = [1, 2, {'a': [3]}]
 97        assert list(flatten(data)) == result
 98
 99    """
100    if isinstance(data, list):
101        for i in data:
102            yield from flatten(i)
103
104    else:
105        yield data

17def get(data: Data,
18        path: Path,
19        default: Data | None = None
20        ) -> Data:
21    """Get data element referenced by path
22
23    Example::
24
25        data = {'a': [1, 2, [3, 4]]}
26        path = ['a', 2, 0]
27        assert get(data, path) == 3
28
29        data = [1, 2, 3]
30        assert get(data, 0) == 1
31        assert get(data, 5) is None
32        assert get(data, 5, default=123) == 123
33
34    """
35    for i in flatten(path):
36        if isinstance(i, str):
37            if not isinstance(data, dict) or i not in data:
38                return default
39            data = data[i]
40
41        elif isinstance(i, int) and not isinstance(i, bool):
42            if not isinstance(data, list):
43                return default
44            try:
45                data = data[i]
46            except IndexError:
47                return default
48
49        else:
50            raise ValueError('invalid path')
51
52    return data

 55def set_(data: Data,
 56         path: Path,
 57         value: Data
 58         ) -> Data:
 59    """Create new data by setting data path element value
 60
 61    Example::
 62
 63        data = [1, {'a': 2, 'b': 3}, 4]
 64        path = [1, 'b']
 65        result = set_(data, path, 5)
 66        assert result == [1, {'a': 2, 'b': 5}, 4]
 67        assert result is not data
 68
 69        data = [1, 2, 3]
 70        result = set_(data, 4, 4)
 71        assert result == [1, 2, 3, None, 4]
 72
 73    """
 74    parents = collections.deque()
 75
 76    for i in flatten(path):
 77        parent = data
 78
 79        if isinstance(i, str):
 80            data = data.get(i) if isinstance(data, dict) else None
 81
 82        elif isinstance(i, int) and not isinstance(i, bool):
 83            try:
 84                data = data[i] if isinstance(data, list) else None
 85            except IndexError:
 86                data = None
 87
 88        else:
 89            raise ValueError('invalid path')
 90
 91        parents.append((parent, i))
 92
 93    while parents:
 94        parent, i = parents.pop()
 95
 96        if isinstance(i, str):
 97            parent = dict(parent) if isinstance(parent, dict) else {}
 98            parent[i] = value
 99
100        elif isinstance(i, int) and not isinstance(i, bool):
101            if not isinstance(parent, list):
102                parent = []
103
104            if i >= len(parent):
105                parent = [*parent,
106                          *itertools.repeat(None, i - len(parent) + 1)]
107
108            elif i < 0 and (-i) > len(parent):
109                parent = [*itertools.repeat(None, (-i) - len(parent)),
110                          *parent]
111
112            else:
113                parent = list(parent)
114
115            parent[i] = value
116
117        else:
118            raise ValueError('invalid path')
119
120        value = parent
121
122    return value

125def remove(data: Data,
126           path: Path
127           ) -> Data:
128    """Create new data by removing part of data referenced by path
129
130    Example::
131
132        data = [1, {'a': 2, 'b': 3}, 4]
133        path = [1, 'b']
134        result = remove(data, path)
135        assert result == [1, {'a': 2}, 4]
136        assert result is not data
137
138        data = [1, 2, 3]
139        result = remove(data, 4)
140        assert result == [1, 2, 3]
141
142    """
143    result = data
144    parents = collections.deque()
145
146    for i in flatten(path):
147        parent = data
148
149        if isinstance(i, str):
150            if not isinstance(data, dict) or i not in data:
151                return result
152            data = data[i]
153
154        elif isinstance(i, int) and not isinstance(i, bool):
155            if not isinstance(data, list):
156                return result
157            try:
158                data = data[i]
159            except IndexError:
160                return result
161
162        else:
163            raise ValueError('invalid path')
164
165        parents.append((parent, i))
166
167    result = None
168
169    while parents:
170        parent, i = parents.pop()
171
172        if isinstance(i, str):
173            parent = dict(parent)
174
175        elif isinstance(i, int) and not isinstance(i, bool):
176            parent = list(parent)
177
178        else:
179            raise ValueError('invalid path')
180
181        if result is None:
182            del parent[i]
183
184        else:
185            parent[i] = result
186
187        result = parent
188
189    return result

192class Storage:
193    """JSON data storage
194
195    Helper class representing observable JSON data state manipulated with
196    path based get/set/remove functions.
197
198    """
199
200    def __init__(self, data: Data = None):
201        self._data = data
202        self._change_cbs = util.CallbackRegistry()
203
204    @property
205    def data(self) -> Data:
206        """Data"""
207        return self._data
208
209    def register_change_cb(self,
210                           cb: Callable[[Data], None]
211                           ) -> util.RegisterCallbackHandle:
212        """Register data change callback"""
213        return self._change_cbs.register(cb)
214
215    def get(self, path: Path, default: Data | None = None):
216        """Get data"""
217        return get(self._data, path, default)
218
219    def set(self, path: Path, value: Data):
220        """Set data"""
221        self._data = set_(self._data, path, value)
222        self._change_cbs.notify(self._data)
223
224    def remove(self, path: Path):
225        """Remove data"""
226        self._data = remove(self._data, path)
227        self._change_cbs.notify(self._data)

17class Format(enum.Enum):
18    """Encoding format"""
19    JSON = 'json'
20    YAML = 'yaml'
21    TOML = 'toml'

24def encode(data: Data,
25           format: Format = Format.JSON,
26           indent: int | None = None
27           ) -> str:
28    """Encode JSON data.
29
30    In case of TOML format, data must be JSON Object.
31
32    Args:
33        data: JSON data
34        format: encoding format
35        indent: indentation size
36
37    """
38    if format == Format.JSON:
39        return json.dumps(data, indent=indent, allow_nan=False)
40
41    if format == Format.YAML:
42        dumper = (yaml.CSafeDumper if hasattr(yaml, 'CSafeDumper')
43                  else yaml.SafeDumper)
44        return str(yaml.dump(data, indent=indent, Dumper=dumper))
45
46    if format == Format.TOML:
47        return tomli_w.dumps(data)
48
49    raise ValueError('unsupported format')

52def decode(data_str: str,
53           format: Format = Format.JSON
54           ) -> Data:
55    """Decode JSON data.
56
57    Args:
58        data_str: encoded JSON data
59        format: encoding format
60
61    """
62    if format == Format.JSON:
63        return json.loads(data_str)
64
65    if format == Format.YAML:
66        loader = (yaml.CSafeLoader if hasattr(yaml, 'CSafeLoader')
67                  else yaml.SafeLoader)
68        return yaml.load(io.StringIO(data_str), Loader=loader)
69
70    if format == Format.TOML:
71        return tomli.loads(data_str)
72
73    raise ValueError('unsupported format')

76def get_file_format(path: pathlib.PurePath) -> Format:
77    """Detect file format based on path suffix"""
78    if path.suffix == '.json':
79        return Format.JSON
80
81    if path.suffix in ('.yaml', '.yml'):
82        return Format.YAML
83
84    if path.suffix == '.toml':
85        return Format.TOML
86
87    raise ValueError('can not determine format from path suffix')

 90def encode_file(data: Data,
 91                path: pathlib.PurePath,
 92                format: Format | None = None,
 93                indent: int | None = 4):
 94    """Encode JSON data to file.
 95
 96    If `format` is ``None``, encoding format is derived from path suffix.
 97
 98    In case of TOML format, data must be JSON Object.
 99
100    Args:
101        data: JSON data
102        path: file path
103        format: encoding format
104        indent: indentation size
105
106    """
107    if format is None:
108        format = get_file_format(path)
109
110    flags = 'w' if format != Format.TOML else 'wb'
111    encoding = 'utf-8' if format != Format.TOML else None
112
113    with open(path, flags, encoding=encoding) as f:
114        encode_stream(data, f, format, indent)

117def decode_file(path: pathlib.PurePath,
118                format: Format | None = None
119                ) -> Data:
120    """Decode JSON data from file.
121
122    If `format` is ``None``, encoding format is derived from path suffix.
123
124    Args:
125        path: file path
126        format: encoding format
127
128    """
129    if format is None:
130        format = get_file_format(path)
131
132    flags = 'r' if format != Format.TOML else 'rb'
133    encoding = 'utf-8' if format != Format.TOML else None
134
135    with open(path, flags, encoding=encoding) as f:
136        return decode_stream(f, format)

139def encode_stream(data: Data,
140                  stream: io.TextIOBase | io.RawIOBase,
141                  format: Format = Format.JSON,
142                  indent: int | None = 4):
143    """Encode JSON data to stream.
144
145    In case of TOML format, data must be JSON Object.
146
147    In case of TOML format, `stream` should be `io.RawIOBase`. For
148    other formats, `io.TextIOBase` is expected.
149
150    Args:
151        data: JSON data
152        stream: output stream
153        format: encoding format
154        indent: indentation size
155
156    """
157    if format == Format.JSON:
158        json.dump(data, stream, indent=indent, allow_nan=False)
159
160    elif format == Format.YAML:
161        dumper = (yaml.CSafeDumper if hasattr(yaml, 'CSafeDumper')
162                  else yaml.SafeDumper)
163        yaml.dump(data, stream, indent=indent, Dumper=dumper,
164                  explicit_start=True, explicit_end=True)
165
166    elif format == Format.TOML:
167        tomli_w.dump(data, stream)
168
169    else:
170        raise ValueError('unsupported format')

173def decode_stream(stream: io.TextIOBase | io.RawIOBase,
174                  format: Format = Format.JSON
175                  ) -> Data:
176    """Decode JSON data from stream.
177
178    In case of TOML format, `stream` should be `io.RawIOBase`. For
179    other formats, `io.TextIOBase` is expected.
180
181    Args:
182        stream: input stream
183        format: encoding format
184
185    """
186    if format == Format.JSON:
187        return json.load(stream)
188
189    if format == Format.YAML:
190        loader = (yaml.CSafeLoader if hasattr(yaml, 'CSafeLoader')
191                  else yaml.SafeLoader)
192        return yaml.load(stream, Loader=loader)
193
194    if format == Format.TOML:
195        return tomli.load(stream)
196
197    raise ValueError('unsupported format')

200def read_conf(path: pathlib.Path | None,
201              default_path: pathlib.Path | None = None,
202              default_suffixes: list[str] = ['.yaml', '.yml', '.toml', '.json'],  # NOQA
203              stdio_path: pathlib.Path | None = pathlib.Path('-')
204              ) -> Data:
205    """Read configuration formated as JSON data"""
206    if stdio_path and path == stdio_path:
207        return decode_stream(sys.stdin)
208
209    if path:
210        return decode_file(path)
211
212    if not default_path:
213        raise Exception('invalid configuration path')
214
215    for suffix in default_suffixes:
216        path = default_path.with_suffix(suffix)
217        if path.exists():
218            break
219
220    return decode_file(path)

 9def diff(src: Data,
10         dst: Data
11         ) -> Data:
12    """Generate JSON Patch diff.
13
14    Example::
15
16        src = [1, {'a': 2}, 3]
17        dst = [1, {'a': 4}, 3]
18        result = diff(src, dst)
19        assert result == [{'op': 'replace', 'path': '/1/a', 'value': 4}]
20
21    """
22    return jsonpatch.JsonPatch.from_diff(src, dst).patch

25def patch(data: Data,
26          diff: Data
27          ) -> Data:
28    """Apply JSON Patch diff.
29
30    Example::
31
32        data = [1, {'a': 2}, 3]
33        d = [{'op': 'replace', 'path': '/1/a', 'value': 4}]
34        result = patch(data, d)
35        assert result == [1, {'a': 4}, 3]
36
37    """
38    return jsonpatch.apply_patch(data, diff)

 18class SchemaRepository:
 19    """JSON Schema repository.
 20
 21    A repository that holds json schemas and enables validation against them.
 22
 23    Repository can be initialized with multiple arguments, which can be
 24    instances of ``pathlib.PurePath``, ``Data`` or ``SchemaRepository``.
 25
 26    If an argument is of type ``pathlib.PurePath``, and path points to file
 27    with a suffix '.json', '.yml' or '.yaml', json serializable data is decoded
 28    from the file. Otherwise, it is assumed that path points to a directory,
 29    which is recursively searched for json and yaml files. All decoded schemas
 30    are added to the repository. If a schema with the same `id` was previously
 31    added, an exception is raised.
 32
 33    If an argument is of type ``Data``, it should be a json serializable data
 34    representation of a JSON schema. If a schema with the same `id` was
 35    previously added, an exception is raised.
 36
 37    If an argument is of type ``SchemaRepository``, its schemas are added to
 38    the new repository. Previously added schemas with the same `id` are
 39    replaced.
 40
 41    """
 42
 43    def __init__(self, *args: typing.Union[pathlib.PurePath,
 44                                           Data,
 45                                           'SchemaRepository']):
 46        self._validators = weakref.WeakValueDictionary()
 47        self._data = collections.defaultdict(dict)
 48
 49        for arg in args:
 50            if isinstance(arg, pathlib.PurePath):
 51                self._load_path(arg)
 52
 53            elif isinstance(arg, SchemaRepository):
 54                self._load_repository(arg)
 55
 56            else:
 57                self._load_schema(arg)
 58
 59    def get_uri_schemes(self) -> Iterable[str]:
 60        """Get URI schemes stored in repository"""
 61        return self._data.keys()
 62
 63    def get_schema_ids(self,
 64                       uri_schemes: Iterable[str] | None = None
 65                       ) -> Iterable[str]:
 66        """Get schema ids stored in repository
 67
 68        If `uri_schemes` is ``None``, all schema ids are returned. Otherwise,
 69        only schema ids that have one of provided URI scheme are returned.
 70
 71        """
 72        if uri_schemes is None:
 73            uri_schemes = self._data.keys()
 74
 75        for uri_scheme in uri_schemes:
 76            schemas = self._data.get(uri_scheme)
 77            if not schemas:
 78                continue
 79
 80            for path in schemas.keys():
 81                yield f'{uri_scheme}://{path}'
 82
 83    def get_schema(self, schema_id: str) -> Data:
 84        """Get stored schema based on schema id"""
 85        uri = urllib.parse.urlparse(schema_id)
 86        path = uri.netloc + uri.path
 87        return self._data[uri.scheme][path]
 88
 89    def validate(self,
 90                 schema_id: str,
 91                 data: Data,
 92                 validator_cls: typing.Type[Validator] = DefaultValidator):
 93        """Validate data against JSON schema.
 94
 95        Args:
 96            schema_id: JSON schema identifier
 97            data: data to be validated
 98            validator_cls: validator implementation
 99
100        Raises:
101            Exception
102
103        """
104        validator = self._validators.get(validator_cls)
105        if validator is None:
106            validator = validator_cls(self)
107            self._validators[validator_cls] = validator
108
109        validator.validate(schema_id, data)
110
111    def to_json(self) -> Data:
112        """Export repository content as json serializable data.
113
114        Entire repository content is exported as json serializable data.
115        New repository can be created from the exported content by using
116        :meth:`SchemaRepository.from_json`.
117
118        """
119        return self._data
120
121    @staticmethod
122    def from_json(data: pathlib.PurePath | Data
123                  ) -> 'SchemaRepository':
124        """Create new repository from content exported as json serializable
125        data.
126
127        Creates a new repository from content of another repository that was
128        exported by using :meth:`SchemaRepository.to_json`.
129
130        Args:
131            data: repository data
132
133        """
134        if isinstance(data, pathlib.PurePath):
135            data = decode_file(data)
136
137        repo = SchemaRepository()
138        repo._data.update(data)
139        return repo
140
141    def _load_path(self, path):
142        json_suffixes = {'.json', '.yaml', '.yml'}
143        paths = ([path] if path.suffix in json_suffixes
144                 else list(itertools.chain.from_iterable(
145                    path.rglob(f'*{i}') for i in json_suffixes)))
146        for path in paths:
147            schema = decode_file(path)
148            self._load_schema(schema)
149
150    def _load_schema(self, schema):
151        if '$schema' in schema:
152            meta_schema_id = urllib.parse.urldefrag(schema['$schema']).url
153            if meta_schema_id not in _meta_schema_ids:
154                schema = dict(schema)
155                del schema['$schema']
156
157        schema_id = schema.get('$id')
158        if not schema_id:
159            schema_id = schema.get('id')
160        if not schema_id:
161            raise Exception('invalid schema id')
162
163        uri = urllib.parse.urlparse(schema_id)
164        path = uri.netloc + uri.path
165        if path in self._data[uri.scheme]:
166            raise Exception(f"duplicate schema id {uri.scheme}://{path}")
167
168        self._data[uri.scheme][path] = schema
169
170    def _load_repository(self, repo):
171        for k, v in repo._data.items():
172            self._data[k].update(v)